Menu
Open source

Options reference

Options define test-run behavior. Most options can be passed in multiple places. If an option is defined in multiple places, k6 chooses the value from the highest order of precedence.

Quick reference of options

Each option has its own detailed reference in a separate section.

OptionDescription
AddressAddress of the REST API server
BatchMax number of simultaneous connections of a http.batch() call
Batch per hostMax number of simultaneous connections of a http.batch() call for a host
Blacklist IPBlacklist IP ranges from being called
Block hostnamesBlock any requests to specific hostnames
Compatibility modeSupport running scripts with different ECMAScript modes
ConfigSpecify the config file in JSON format to read the options values
Console outputRedirects logs logged by console methods to the provided output file
Discard response bodiesSpecify whether response bodies should be discarded
DNSConfigure DNS resolution behavior
DurationA string specifying the total duration of the test run; together with the vus option, it’s a shortcut for a single scenario with a constant VUs executor
Execution segmentLimit execution to a segment of the total test
Exit on runningExits when test reaches the running status
Cloud optionsAn object used to set configuration options for cloud parameters.
HostsAn object with overrides to DNS resolution
HTTP debugLog all HTTP requests and responses
Include system Env varsPass the real system environment variables to the runtime
Insecure skip TLS verifyA boolean specifying whether k6 should ignore TLS verifications for connections established from code
IterationsA number specifying a fixed number of iterations to execute of the script; together with the vus option, it’s a shortcut for a single scenario with a shared iterations executor
LingerA boolean specifying whether k6 should linger around after test run completion
Local IPsA list of local IPs, IP ranges, and CIDRs from which VUs will make requests
Log outputConfiguration about where logs from k6 should be send
LogFormatSpecify the format of the log output
Max redirectsThe maximum number of HTTP redirects that k6 will follow
Minimum iteration durationSpecify the minimum duration for every single execution
No colorA boolean specifying whether colored output is disabled
No connection reuseA boolean specifying whether k6 should disable keep-alive connections
No cookies resetThis disables resetting the cookie jar after each VU iteration
No summarydisables the end-of-test summary
No setupA boolean specifying whether setup() function should be run
No teardownA boolean specifying whether teardown() function should be run
No thresholdsDisables threshold execution
No usage reportA boolean specifying whether k6 should send a usage report
No VU connection reuseA boolean specifying whether k6 should reuse TCP connections
PausedA boolean specifying whether the test should start in a paused state
Profiling EnabledEnables profiling endpoints
QuietA boolean specifying whether to show the progress update in the console or not
Results outputSpecify the results output
RPSThe maximum number of requests to make per second globally (discouraged, use arrival-rate executors instead)
ScenariosDefine advanced execution scenarios
Setup timeoutSpecify how long the setup() function is allow to run before it’s terminated
Show logsA boolean specifying whether the cloud logs are printed out to the terminal
StagesA list of objects that specify the target number of VUs to ramp up or down; shortcut option for a single scenario with a ramping VUs executor
Supply environment variableAdd/override environment variable with VAR=value
System tagsSpecify which System Tags will be in the collected metrics
Summary exportOutput the end-of-test summary report to a JSON file (discouraged, use handleSummary() instead)
Summary trend statsDefine stats for trend metrics in the end-of-test summary
Summary time unitTime unit to be used for all time values in the end-of-test summary
TagsSpecify tags that should be set test-wide across all metrics
Teardown timeoutSpecify how long the teardown() function is allowed to run before it’s terminated
ThresholdsConfigure under what conditions a test is successful or not
ThrowA boolean specifying whether to throw errors on failed HTTP requests
TLS authA list of TLS client certificate configuration objects
TLS cipher suitesA list of cipher suites allowed to be used by in SSL/TLS interactions with a server
TLS versionString or object representing the only SSL/TLS version allowed
Traces outputConfiguration about where traces from k6 should be sent
User agentA string specifying the User-Agent header when sending HTTP requests
VerboseA boolean specifying whether verbose logging is enabled
VUsA number specifying the number of VUs to run concurrently

The following sections detail all available options that you can be specify within a script.

It also documents the equivalent command line flag, environment variables or option when executing k6 run ... and k6 cloud run ..., which you can use to override options specified in the code.

Address

Address of the API server. When executing scripts with k6 run an HTTP server with a REST API is spun up, which can be used to control some of the parameters of the test execution. By default, the server listens on localhost:6565. Read more on k6 REST API.

EnvCLICode / Config fileDefault
N/A--address, -aN/Alocalhost:6565
bash
$ k6 run --address "localhost:3000" script.js

Batch

The maximum number of simultaneous/parallel connections in total that an http.batch() call in a VU can make. If you have a batch() call that you’ve given 20 URLs to and --batch is set to 15, then the VU will make 15 requests right away in parallel and queue the rest, executing them as soon as a previous request is done and a slot opens. Available in both the k6 run and the k6 cloud run commands

EnvCLICode / Config fileDefault
K6_BATCH--batchbatch20
JavaScript
export const options = {
  batch: 15,
};

Batch per host

The maximum number of simultaneous/parallel connections for the same hostname that an http.batch() call in a VU can make. If you have a batch() call that you’ve given 20 URLs to the same hostname and --batch-per-host is set to 5, then the VU will make 5 requests right away in parallel and queue the rest, executing them as soon as a previous request is done and a slot opens. This will not run more request in parallel then the value of batch. Available in both the k6 run and the k6 cloud run commands

EnvCLICode / Config fileDefault
K6_BATCH_PER_HOST--batch-per-hostbatchPerHost6
JavaScript
export const options = {
  batchPerHost: 5,
};

Blacklist IP

Blacklist IP ranges from being called. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_BLACKLIST_IPS--blacklist-ipblacklistIPsnull
JavaScript
export const options = {
  blacklistIPs: ['10.0.0.0/8'],
};

Block hostnames

Blocks hostnames based on a list of glob match strings. The pattern matching string can have a single * at the beginning such as *.example.com that will match anything before that such as test.example.com and test.test.example.com. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_BLOCK_HOSTNAMES--block-hostnamesblockHostnamesnull
JavaScript
export const options = {
  blockHostnames: ['test.k6.io', '*.example.com'],
};
bash
$ k6 run --block-hostnames="test.k6.io,*.example.com" script.js

Compatibility mode

Support running scripts with different ECMAScript compatibility modes.

Read about the different modes on the JavaScript Compatibility Mode documentation.

EnvCLICode / Config fileDefault
K6_COMPATIBILITY_MODE--compatibility-modeN/A"extended"
bash
$ k6 run --compatibility-mode=base script.js

Config

Specify the config file in JSON format. If the config file is not specified, k6 looks for config.json in the loadimpact/k6 directory inside the regular directory for configuration files on the operating system. Default config locations on different operating systems are as follows:

OSDefault Config Path
Unix-based${HOME}/.config/loadimpact/k6/config.json
macOS${HOME}/Library/Application Support/loadimpact/k6/config.json
Windows%AppData%/loadimpact/k6/config.json

Available in k6 run and k6 cloud run commands:

EnvCLICode / Config fileDefault
N/A--config <path>, -c <path>N/Anull

Note

When running tests in k6 Cloud and using a non-default config.json file, specify the cloud token inside your config file to authenticate.

Console output

Redirects logs logged by console methods to the provided output file. Available in k6 cloud run and k6 run commands.

EnvCLICode / Config fileDefault
K6_CONSOLE_OUTPUT--console-outputN/Anull
bash
$ k6 run --console-output "loadtest.log" script.js

Discard response bodies

Specify if response bodies should be discarded by changing the default value of responseType to none for all HTTP requests. Highly recommended to be set to true and then only for the requests where the response body is needed for scripting to set responseType to text or binary. Lessens the amount of memory required and the amount of GC - reducing the load on the testing machine, and probably producing more reliable test results.

EnvCLICode / Config fileDefault
K6_DISCARD_RESPONSE_BODIES--discard-response-bodiesdiscardResponseBodiesfalse
JavaScript
export const options = {
  discardResponseBodies: true,
};

DNS

This is a composite option that provides control of DNS resolution behavior with configuration for cache expiration (TTL), IP selection strategy and IP version preference. The TTL field in the DNS record is currently not read by k6, so the ttl option allows manual control over this behavior, albeit as a fixed value for the duration of the test run.

Note that DNS resolution is done only on new HTTP connections, and by default k6 will try to reuse connections if HTTP keep-alive is supported. To force a certain DNS behavior consider enabling the noConnectionReuse option in your tests.

EnvCLICode / Config fileDefault
K6_DNS--dnsdnsttl=5m,select=random,policy=preferIPv4

Possible ttl values are:

  • 0: no caching at all - each request will trigger a new DNS lookup.
  • inf: cache any resolved IPs for the duration of the test run.
  • any time duration like 60s, 5m30s, 10m, 2h, etc.; if no unit is specified (e.g. ttl=3000), k6 assumes milliseconds.

Possible select values are:

  • first: always pick the first resolved IP.
  • random: pick a random IP for every new connection.
  • roundRobin: iterate sequentially over the resolved IPs.

Possible policy values are:

  • preferIPv4: use IPv4 addresses if available, otherwise fall back to IPv6.
  • preferIPv6: use IPv6 addresses if available, otherwise fall back to IPv4.
  • onlyIPv4: only use IPv4 addresses, ignore any IPv6 ones.
  • onlyIPv6: only use IPv6 addresses, ignore any IPv4 ones.
  • any: no preference, use all addresses.

Here are some configuration examples:

bash
K6_DNS="ttl=5m,select=random,policy=preferIPv4" k6 cloud run script.js
JavaScript
export const options = {
  dns: {
    ttl: '1m',
    select: 'roundRobin',
    policy: 'any',
  },
};

Duration

A string specifying the total duration a test run should be run for. During this time each VU will execute the script in a loop. Available in k6 run and k6 cloud run commands.

Together with the vus option, duration is a shortcut for a single scenario with a constant VUs executor.

EnvCLICode / Config fileDefault
K6_DURATION--duration, -ddurationnull
JavaScript
export const options = {
  vus: 100,
  duration: '3m',
};

Cloud options

An object used to set configuration options for Grafana Cloud k6. For more information about available parameters, refer to Cloud options.

EnvCLICode / Config fileDefault
N/AN/Acloudnull

This is an example of how to specify the test name (test runs/executions with the same name will be logically grouped for trending and comparison) when streaming results to Grafana Cloud k6.

JavaScript
export const options = {
  cloud: {
    name: 'My test name',
  },
};

Previously, the cloud object was known as ext.loadimpact.

Execution segment

These options specify how to partition the test run and which segment to run. If defined, k6 will scale the number of VUs and iterations to be run for that segment, which is useful in distributed execution. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
N/A--execution-segmentexecutionSegment"0:1"
N/A--execution-segment-sequenceexecutionSegmentSequence"0,1"

For example, to run 25% of a test, you would specify --execution-segment '25%', which would be equivalent to --execution-segment '0:1/4', i.e. run the first 1/4 of the test. To ensure that each instance executes a specific segment, also specify the full segment sequence, e.g. --execution-segment-sequence '0,1/4,1/2,1'. This way one instance could run with --execution-segment '0:1/4', another with --execution-segment '1/4:1/2', etc. and there would be no overlap between them.

Exit on running

A boolean, specifying whether the script should exit once the test status reaches running. When running scripts with k6 cloud run by default scripts will run until the test reaches a finalized status. This could be problematic in certain environments (think of Continuous Integration and Delivery pipelines), since you’d need to wait until the test ends up in a finalized state.

With this option, you can exit early and let the script run in the background. Available in k6 cloud run command.

EnvCLICode / Config fileDefault
K6_EXIT_ON_RUNNING--exit-on-runningN/Afalse
bash
$ k6 cloud run --exit-on-running script.js

Hosts

An object with overrides to DNS resolution, similar to what you can do with /etc/hosts on Linux/Unix or C:\Windows\System32\drivers\etc\hosts on Windows. For instance, you could set up an override which routes all requests for test.k6.io to 1.2.3.4.

k6 also supports ways to narrow or widen the scope of your redirects:

  • You can redirect only from or to certain ports.
  • Starting from v0.42.0, you can use an asterisk (*) as a wild card at the start of the host name to avoid repetition. For example, *.k6.io would apply the override for all subdomains of k6.io.

Note

This does not modify the actual HTTP Host header, but rather where it will be routed.
EnvCLICode / Config fileDefault
N/AN/Ahostsnull
JavaScript
export const options = {
  hosts: {
    'test.k6.io': '1.2.3.4',
    'test.k6.io:443': '1.2.3.4:8443',
    '*.grafana.com': '1.2.3.4',
  },
};

The preceding code will redirect requests made to test.k6.io to 1.2.3.4, keeping the same port. If the request is done to port 443, it will redirect it to port 8443 instead. It will also redirect requests to any subdomain of grafana.com to 1.2.3.4.

HTTP debug

Log all HTTP requests and responses. Excludes body by default, to include body use --http-debug=full. Available in k6 run and k6 cloud run commands.

Read more here.

EnvCLICode / Config fileDefault
K6_HTTP_DEBUG--http-debug,
--http-debug=full
httpDebugfalse
JavaScript
export const options = {
  httpDebug: 'full',
};

Include system env vars

Pass the real system environment variables to the runtime. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
N/A--include-system-env-varsN/Atrue for k6 run, but false for all other commands to prevent inadvertent sensitive data leaks.
bash
$ k6 run --include-system-env-vars ~/script.js

Insecure skip TLS verify

A boolean, true or false. When this option is enabled (set to true), all of the verifications that would otherwise be done to establish trust in a server provided TLS certificate will be ignored. This only applies to connections created from code, such as HTTP requests. Available in k6 run and k6 cloud run commands

EnvCLICode / Config fileDefault
K6_INSECURE_SKIP_TLS_VERIFY--insecure-skip-tls-verifyinsecureSkipTLSVerifyfalse
JavaScript
export const options = {
  insecureSkipTLSVerify: true,
};

Iterations

An integer value, specifying the total number of iterations of the default function to execute in the test run, as opposed to specifying a duration of time during which the script would run in a loop. Available both in the k6 run and k6 cloud run commands.

Together with the vus option, iterations is a shortcut for a single scenario with a shared iterations executor.

By default, the maximum duration of a shared-iterations scenario is 10 minutes. You can adjust that time via the maxDuration option of the scenario, or by also specifying the duration global shortcut option.

Note that iterations aren’t fairly distributed with this option, and a VU that executes faster will complete more iterations than others. Each VU will try to complete as many iterations as possible, “stealing” them from the total number of iterations for the test. So, depending on iteration times, some VUs may complete more iterations than others. If you want guarantees that every VU will complete a specific, fixed number of iterations, use the per-VU iterations executor.

EnvCLICode / Config fileDefault
K6_ITERATIONS--iterations, -iiterations1
JavaScript
export const options = {
  vus: 5,
  iterations: 10,
};

Or, to run 10 VUs 10 times each:

JavaScript
export const options = {
  vus: 10,
  iterations: 100,
};

Linger

A boolean, true or false, specifying whether the k6 process should linger around after test run completion. Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_LINGER--linger, -llingerfalse
JavaScript
export const options = {
  linger: true,
};

Local IPs

A list of IPs, IP ranges and CIDRs from which VUs will make requests. The IPs will be sequentially given out to VUs. This option doesn’t change anything on the OS level so the IPs need to be already configured on the OS level for k6 to use them. Also IPv4 CIDRs with more than 2 IPs don’t include the first and last IP as they are reserved for referring to the network itself and the broadcast address respectively.

This option can be used for splitting the network traffic from k6 between multiple network cards, thus potentially increasing the available network throughput. For example, if you have 2 NICs, you can run k6 with --local-ips="<IP-from-first-NIC>,<IP-from-second-NIC>" to balance the traffic equally between them - half of the VUs will use the first IP and the other half will use the second. This can scale to any number of NICs, and you can repeat some local IPs to give them more traffic. For example, --local-ips="<IP1>,<IP2>,<IP3>,<IP3>" will split VUs between 3 different source IPs in a 25%:25%:50% ratio.

Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_LOCAL_IPS--local-ipsN/AN/A
bash
$ k6 run --local-ips=192.168.20.12-192.168.20.15,192.168.10.0/27 script.js

Log output

This option specifies where to send logs to and another configuration connected to it. Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_LOG_OUTPUT--log-outputN/Astderr
bash
$ k6 run --log-output=stdout script.js

Possible values are:

  • none - disable
  • stdout - send to the standard output
  • stderr - send to the standard error output (this is the default)
  • loki - send logs to a loki server
  • file - write logs to a file

Loki

Use the log-output option to configure Loki as follows. For additional instructions and a step-by-step guide, check out the Loki tutorial.

bash
$ k6 run --log-output=loki=http://127.0.0.1:3100/loki/api/v1/push,label.something=else,label.foo=bar,limit=32,level=info,pushPeriod=5m32s,msgMaxSize=1231 script.js

Where all but the url in the beginning are not required. The possible keys with their meanings and default values:

keydescriptiondefault value
nothingthe endpoint to which to send logshttp://127.0.0.1:3100/loki/api/v1/push
allowedLabelsif set k6 will send only the provided labels as such and all others will be appended to the message in the form key=value. The value of the option is in the form [label1,label2]N/A
label.labelNameadds an additional label with the provided key and value to each messageN/A
header.headerNameadds an additional HTTP header with the provided header name and value to each HTTP request made to LokiN/A
limitthe limit of message per pushPeriod, an additional log is send when the limit is reached, logging how many logs were dropped100
levelthe minimal level of a message so it’s send to lokiall
pushPeriodat what period to send log lines1s
profilewhether to print some info about performance of the sending to lokifalse
msgMaxSizehow many symbols can there be at most in a message. Messages bigger will miss the middle of the message with an additional few characters explaining how many characters were dropped.1048576

File

The file can be configured as below, where an explicit file path is required:

bash
$ k6 run --log-output=file=./k6.log script.js

A valid file path is the unique mandatory field, the other optional fields listed below:

keydescriptiondefault value
levelthe minimal level of a message to write out of (in ascending order): trace, debug, info, warning, error, fatal, panictrace

LogFormat

A value specifying the log format. By default, k6 includes extra debug information like date and log level. The other options available are:

  • json: print all the debug information in JSON format.

  • raw: print only the log message.

EnvCLICode / Config fileDefault
K6_LOG_FORMAT--log-format, -fN/A
bash
$ k6 run --log-format raw test.js

Max redirects

The maximum number of HTTP redirects that k6 will follow before giving up on a request and erroring out. Available in both the k6 run and the k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_MAX_REDIRECTS--max-redirectsmaxRedirects10
JavaScript
export const options = {
  maxRedirects: 10,
};

Minimum iteration duration

Specifies the minimum duration of every single execution (i.e. iteration) of the default function. Any iterations that are shorter than this value will cause that VU to sleep for the remainder of the time until the specified minimum duration is reached.

EnvCLICode / Config fileDefault
K6_MIN_ITERATION_DURATION--min-iteration-durationminIterationDuration0 (disabled)
JavaScript
export const options = {
  minIterationDuration: '10s',
};

No color

A boolean specifying whether colored output is disabled. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
N/A--no-colorN/Afalse
bash
$ k6 run --no-color script.js

No connection reuse

A boolean, true or false, specifying whether k6 should disable keep-alive connections. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_NO_CONNECTION_REUSE--no-connection-reusenoConnectionReusefalse
JavaScript
export const options = {
  noConnectionReuse: true,
};

No cookies reset

This disables the default behavior of resetting the cookie jar after each VU iteration. If it’s enabled, saved cookies will be persisted across VU iterations.

EnvCLICode / Config fileDefault
K6_NO_COOKIES_RESETN/AnoCookiesResetfalse
JavaScript
export const options = {
  noCookiesReset: true,
};

No summary

Disables end-of-test summary generation, including calls to handleSummary() and --summary-export.

Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_NO_SUMMARY--no-summaryN/Afalse
bash
$ k6 run --no-summary ~/script.js

No setup

A boolean specifying whether setup() function should be run. Available in k6 cloud run and k6 run commands.

EnvCLICode / Config fileDefault
K6_NO_SETUP--no-setupN/Afalse
bash
$ k6 run --no-setup script.js

No teardown

A boolean specifying whether teardown() function should be run. Available in k6 cloud run and k6 run commands.

EnvCLICode / Config fileDefault
K6_NO_TEARDOWN--no-teardownN/Afalse
bash
$ k6 run --no-teardown script.js

No thresholds

Disables threshold execution. Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_NO_THRESHOLDS--no-thresholdsN/Afalse
bash
$ k6 run --no-thresholds ~/script.js

No usage report

A boolean, true or false. By default, k6 sends a usage report each time it is run, so that we can track how often people use it. If this option is set to true, no usage report will be made. To learn more, have a look at the Usage reports documentation. Available in k6 run commands.

EnvCLIConfig fileDefault
K6_NO_USAGE_REPORT--no-usage-reportnoUsageReport*false
bash
$ k6 run --no-usage-report ~/script.js

* Note that this option is not supported in the exported script options, but can be specified in a configuration file.

No VU connection reuse

A boolean, true or false, specifying whether k6 should reuse TCP connections between iterations of a VU. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_NO_VU_CONNECTION_REUSE--no-vu-connection-reusenoVUConnectionReusefalse
JavaScript
export const options = {
  noVUConnectionReuse: true,
};

Paused

A boolean, true or false, specifying whether the test should start in a paused state. To resume a paused state you’d use the k6 resume command. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
K6_PAUSED--paused, -ppausedfalse
JavaScript
export const options = {
  paused: true,
};

Profiling Enabled

Enables pprof profiling endpoints under the k6’s REST API address. These endpoints help debug and profile k6 itself. k6’s REST API should be enabled as well.

EnvCLICode / Config fileDefault
K6_PROFILING_ENABLED--profiling-enabledN/Ahttp://localhost:6565/debug/pprof/
bash
$ k6 run --profiling-enabled script.js

Quiet

A boolean, true or false, that disables the progress update bar on the console output. Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
N/A--quiet, -qN/Afalse
bash
$ k6 run script.js -d 20s --quiet

Results output

Specify the results output. For information on all available outputs and how to configure them, refer to Results output.

EnvCLICode / Config fileDefault
K6_OUT--out, -oN/Anull
bash
$ k6 run --out influxdb=http://localhost:8086/k6 script.js

RPS

The maximum number of requests to make per second, in total across all VUs. Available in k6 run and k6 cloud run commands.

Caution

This option is discouraged.

The --rps option has caveats and is difficult to use correctly.

For example, in the cloud or distributed execution, this option affects every k6 instance independently. That is, it is not sharded like VUs are.

We strongly recommend the arrival-rate executors to simulate constant RPS instead of this option.
EnvCLICode / Config fileDefault
K6_RPS--rpsrps0 (unlimited)
JavaScript
export const options = {
  rps: 500,
};

Considerations when running in the cloud

The option is set per load generator which means that the value you set in the options object of your test script will be multiplied by the number of load generators your test run is using. At the moment we are hosting 300 VUs per load generator instance. In practice that means that if you set the option for 100 rps, and run a test with 1000 VUs, you will spin up 4 load gen instances and effective rps limit of your test run will be 400

Scenarios

Define one or more execution patterns, with various VU and iteration scheduling settings, running different exported functions (besides default!), using different environment variables, tags, and more.

See the Scenarios article for details and more examples.

Available in k6 run and k6 cloud run commands.

EnvCLICode / Config fileDefault
N/AN/Ascenariosnull
JavaScript
export const options = {
  scenarios: {
    my_api_scenario: {
      // arbitrary scenario name
      executor: 'ramping-vus',
      startVUs: 0,
      stages: [
        { duration: '5s', target: 100 },
        { duration: '5s', target: 0 },
      ],
      gracefulRampDown: '10s',
      env: { MYVAR: 'example' },
      tags: { my_tag: 'example' },
    },
  },
};

Setup timeout

Specify how long the setup() function can run before it’s terminated and the test fails. If defined, the value must be positive.

EnvCLICode / Config fileDefault
K6_SETUP_TIMEOUTN/AsetupTimeout"60s"
JavaScript
export const options = {
  setupTimeout: '30s',
};

Show logs

A boolean specifying whether the cloud logs are printed out to the terminal. Available in k6 cloud run command.

EnvCLICode / Config fileDefault
N/A--show-logsN/Atrue
bash
$ k6 cloud run --show-logs=false script.js

Stages

A list of VU { target: ..., duration: ... } objects that specify the target number of VUs to ramp up or down to for a specific period. Available in k6 run and k6 cloud run commands.

It is a shortcut option for a single scenario with a ramping VUs executor. If used together with the VUs option, the vus value is used as the startVUs option of the executor.

EnvCLICode / Config fileDefault
K6_STAGES--stage <duration>:<target>, -s <duration>:<target>stagesBased on vus and duration.
JavaScript
// The following config would have k6 ramping up from 1 to 10 VUs for 3 minutes,
// then staying flat at 10 VUs for 5 minutes, then ramping up from 10 to 35 VUs
// over the next 10 minutes before finally ramping down to 0 VUs for another
// 3 minutes.

export const options = {
  stages: [
    { duration: '3m', target: 10 },
    { duration: '5m', target: 10 },
    { duration: '10m', target: 35 },
    { duration: '3m', target: 0 },
  ],
};
bash
$ k6 run --stage 5s:10,5m:20,10s:5 script.js

# or...

$ K6_STAGES="5s:10,5m:20,10s:5" k6 run script.js
windows
C:\k6> k6 run --stage 5s:10,5m:20,10s:5 script.js

# or...

C:\k6> set "K6_STAGES=5s:10,5m:20,10s:5" && k6 run script.js
powershell
C:\k6> k6 run --stage 5s:10,5m:20,10s:5 script.js

# or...

C:\k6> $env:K6_STAGES="5s:10,5m:20,10s:5"; k6 run script.js

Summary export

Save the end-of-test summary report to a JSON file that includes data for all test metrics, checks and thresholds. This is useful to get the aggregated test results in a machine-readable format, for integration with dashboards, external alerts, CI pipelines, etc.

While this feature is not deprecated yet, we now discourage it. For a better, more flexible JSON export, as well as export of the summary data to different formats (e.g. JUnit/XUnit/etc. XML, HTML, .txt) and complete summary customization, use the handleSummary() function.

Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_SUMMARY_EXPORT--summary-export <filename>N/Anull
bash
$ k6 run --summary-export export.json script.js

# or...

$ K6_SUMMARY_EXPORT="export.json" k6 run script.js
windows
C:\k6> k6 run --summary-export export.json script.js

# or...

C:\k6> set "K6_SUMMARY_EXPORT=export.json" && k6 run script.js
powershell
C:\k6> k6 run --summary-export export.json script.js

# or...

C:\k6> $env:K6_SUMMARY_EXPORT="export.json"; k6 run script.js

See an example file on the Results Output page.

Supply environment variables

Add/override an environment variable with VAR=valuein a k6 script. Available in k6 run and k6 cloud commands.

To make the system environment variables available in the k6 script via __ENV, use the --include-system-env-vars option.

Note

The -e flag does not configure options.

This flag just provides variables to the script, which the script can use or ignore. For example, -e K6_ITERATIONS=120 does not configure the script iterations.

Compare this behavior with K6_ITERATIONS=120 k6 run script.js, which does set iterations.

EnvCLICode / Config fileDefault
N/A--env, -eN/Anull
bash
$ k6 run -e FOO=bar ~/script.js

System tags

Specify which system tags will be in the collected metrics. Some collectors like the cloud one may require that certain system tags be used. You can specify the tags as an array from the JS scripts or as a comma-separated list via the CLI. Available in k6 run and k6 cloud commands

EnvCLICode / Config fileDefault
K6_SYSTEM_TAGS--system-tagssystemTagsproto,subproto,status,method,url,name,group,check,error,error_code,tls_version,scenario,service,expected_response
JavaScript
export const options = {
  systemTags: ['status', 'method', 'url'],
};

Summary time unit

Define which time unit will be used for all time values in the end-of-test summary. Possible values are s (seconds), ms (milliseconds) and us (microseconds). If no value is specified, k6 will use mixed time units, choosing the most appropriate unit for each value.

EnvCLICode / Config fileDefault
K6_SUMMARY_TIME_UNIT--summary-time-unitsummaryTimeUnitnull
JavaScript
export const options = {
  summaryTimeUnit: 'ms',
};

Summary trend stats

Define which stats for Trend metrics (e.g. response times, group/iteration durations, etc.) will be shown in the end-of-test summary. Possible values include avg (average), med (median), min, max, count, as well as arbitrary percentile values (e.g. p(95), p(99), p(99.99), etc.).

For further summary customization and exporting the summary in various formats (e.g. JSON, JUnit/XUnit/etc. XML, HTML, .txt, etc.), use the handleSummary() function.

EnvCLICode / Config fileDefault
K6_SUMMARY_TREND_STATS--summary-trend-statssummaryTrendStatsavg,min,med,max,p(90),p(95)
JavaScript
export const options = {
  summaryTrendStats: ['avg', 'min', 'med', 'max', 'p(95)', 'p(99)', 'p(99.99)', 'count'],
};
bash
$ k6 run --summary-trend-stats="avg,min,med,max,p(90),p(99.9),p(99.99),count" ./script.js

Tags

Specify tags that should be set test wide across all metrics. If a tag with the same name has been specified on a request, check or custom metrics it will have precedence over a test wide tag. Available in k6 run and k6 cloud commands.

EnvCLICode / Config fileDefault
N/A--tag NAME=VALUEtagsnull
JavaScript
export const options = {
  tags: {
    name: 'value',
  },
};

Teardown timeout

Specify how long the teardown() function is allowed to run before it’s terminated and the test fails.

EnvCLICode / Config fileDefault
K6_TEARDOWN_TIMEOUTN/AteardownTimeout"60s"
JavaScript
export const options = {
  teardownTimeout: '30s',
};

Thresholds

A collection of threshold specifications to configure under what condition(s) a test is considered successful or not, when it has passed or failed, based on metric data. To learn more, have a look at the Thresholds documentation. Available in k6 run commands.

EnvCLICode / Config fileDefault
N/AN/Athresholdsnull
JavaScript
export const options = {
  thresholds: {
    'http_req_duration': ['avg<100', 'p(95)<200'],
    'http_req_connecting{cdnAsset:true}': ['p(95)<100'],
  },
};

Throw

A boolean, true or false, specifying whether k6 should throw exceptions when certain errors occur, or if it should just log them with a warning. Behaviors that currently depend on this option:

Available in k6 run and k6 cloud commands.

EnvCLICode / Config fileDefault
K6_THROW--throw, -wthrowfalse
JavaScript
export const options = {
  throw: true,
};

TLS auth

A list of TLS client certificate configuration objects. domains and password are optional, but cert and key are required.

EnvCLICode / Config fileDefault
N/AN/AtlsAuthnull
JavaScript
export const options = {
  tlsAuth: [
    {
      domains: ['example.com'],
      cert: open('mycert.pem'),
      key: open('mycert-key.pem'),
      password: 'mycert-passphrase',
    },
  ],
};

TLS cipher suites

A list of cipher suites allowed to be used by in SSL/TLS interactions with a server. For a full listing of available ciphers go here.

Caution

Due to limitations in the underlying go implementation, changing the ciphers for TLS 1.3 is not supported and will do nothing.
EnvCLICode / Config fileDefault
N/AN/AtlsCipherSuitesnull (Allow all suites)
JavaScript
export const options = {
  tlsCipherSuites: ['TLS_RSA_WITH_RC4_128_SHA', 'TLS_RSA_WITH_AES_128_GCM_SHA256'],
};

TLS version

Either a string representing the only SSL/TLS version allowed to be used in interactions with a server, or an object specifying the “min” and “max” versions allowed to be used.

EnvCLICode / Config fileDefault
N/AN/AtlsVersionnull (Allow all versions)
JavaScript
export const options = {
  tlsVersion: 'tls1.2',
};
JavaScript
export const options = {
  tlsVersion: {
    min: 'ssl3.0',
    max: 'tls1.2',
  },
};

Traces output

This option specifies where to send traces to. Available in the k6 run command.

EnvCLICode / Config fileDefault
K6_TRACES_OUTPUT--traces-outputN/Anone
bash
$ k6 run --traces-output=otel script.js

Possible values are:

  • none - disable
  • otel - send traces to an OTEL compatible backend

OTEL

Use the traces-output option to configure Open Telemetry compatible output as follows.

bash
$ k6 run --traces-output=otel=http://127.0.0.1:4318,proto=http,header.AdditionalHeader=example script.js

Where none of the options are required. The possible keys with their meanings and default values:

keydescriptiondefault value
nothingthe endpoint to which send traces tohttp://127.0.0.1:4317
protothe protocol to use when connecting with the traces backendgrpc
header.headerNameadds an additional HTTP header with the provided header name and value to each HTTP request made to the traces backendN/A

Upload only (deprecated)

Caution

The “Upload only” option is deprecated and will be removed in a future release. Use the k6 cloud upload command instead.

A boolean specifying whether the test should just be uploaded to the cloud, but not run it. Available in k6 cloud command.

This would be useful if you would like to update a given test and run it later. For example, updating test scripts of a scheduled test from the CI pipelines.

EnvCLICode / Config fileDefault
K6_CLOUD_UPLOAD_ONLY--upload-onlyN/Afalse
bash
$ k6 cloud run --upload-only script.js

User agent

A string specifying the user-agent string to use in User-Agent headers when sending HTTP requests. If you pass an empty string, no User-Agent header is sent. Available in k6 run and k6 cloud commands

EnvCLICode / Config fileDefault
K6_USER_AGENT--user-agentuserAgentk6/0.27.0 (https://k6.io/) (depending on the version you’re using)`
JavaScript
export const options = {
  userAgent: 'MyK6UserAgentString/1.0',
};

Verbose

A boolean specifying whether verbose logging is enabled. Available in k6 run and k6 cloud commands.

EnvCLICode / Config fileDefault
N/A--verbose, -vN/Afalse
bash
$ k6 run --verbose script.js

VUs

An integer value specifying the number of VUs to run concurrently, used together with the iterations or duration options. If you’d like more control look at the stages option or scenarios.

Available in k6 run and k6 cloud commands.

EnvCLICode / Config fileDefault
K6_VUS--vus, -uvus1
JavaScript
export const options = {
  vus: 10,
  duration: '1h',
};