Skip to main content
Fireworks AI logo

Is Fireworks AI Down?

No — Fireworks AI is up
Fireworks AI's status page
Operational · no open incidents
FlareWarden checks
Website reachable · 8 of 8 regions · 477 ms

Right now

The verdict is the worse of these two. How we decide

Fireworks AI's status page

status.fireworks.ai · read · checked every 3 minutes

Operational
Fireworks AI's components, as its status page reports them
ComponentStatus
MiniMax M3.1 FlashOperational
DeepSeek V4.1 Flash 1017Operational
GLM 5.3 FlashOperational
GLM 5.3Operational
GLM 5.3 FastOperational
Ember-1Operational
Kimi K3 FastOperational
Kimi K3Operational
Qwen 3.8 MaxOperational
Deepseek V4.1 FlashOperational
MiniMax M3Operational
InklingOperational
1 more, all operational · show all 13
All of Fireworks AI's components
ComponentStatus
MiniMax M3.1 FlashOperational
DeepSeek V4.1 Flash 1017Operational
GLM 5.3 FlashOperational
GLM 5.3Operational
GLM 5.3 FastOperational
Ember-1Operational
Kimi K3 FastOperational
Kimi K3Operational
Qwen 3.8 MaxOperational
Deepseek V4.1 FlashOperational
MiniMax M3Operational
InklingOperational
GPT-OSS 120BOperational
9 locations All operational
Open their status page

FlareWarden checks

fireworks.ai from 8 regions · when you open this page (at most every 30 minutes), otherwise every 6 hours

Reachable
Regions up
8 of 8
Response
477 ms
Checked

By region

  • Amsterdam reachable, 318 ms
  • Dallas reachable, 326 ms
  • Paris reachable, 254 ms
  • San Jose reachable, 370 ms
  • Singapore reachable, 1047 ms
  • Stockholm reachable, 308 ms
  • Sydney reachable, 833 ms
  • Toronto reachable, 366 ms
Timing by region
Connection timing for each region that loaded the page
RegionDNSTCPTLSTTFB
Amsterdam26 ms6 ms20 ms288 ms
Dallas56 ms0 ms53 ms277 ms
Paris16 ms9 ms16 ms215 ms
San Jose76 ms1 ms7 ms314 ms
Singapore16 ms1 ms7 ms532 ms
Stockholm71 ms1 ms10 ms245 ms
Sydney5 ms1 ms6 ms411 ms
Toronto41 ms16 ms37 ms295 ms
We check that fireworks.ai loads, not every feature behind it. Test it yourself

History

Incidents Fireworks AI reported

Since Oct 22, 3 days · from status.fireworks.ai

Incidents
7
Major or critical
0
Median to resolve
29m

  1. Service Degradation for GLM 5.3

    minor ·

    lasted 3h 5m

  2. Service Degradation for DeepSeek V4.1 Flash 1017

    minor ·

    lasted 1m

  3. Service Degradation for DeepSeek V4.1 Flash 1017

    minor ·

    lasted 24m

  4. Service Degradation for Deepseek V4.1 Flash

    minor ·

    lasted 55m

  5. Service Degradation for DeepSeek V4.1 Flash 1017

    minor ·

    lasted 1h 1m

  6. Service Degradation for Deepseek V4.1 Flash

    minor ·

    lasted 29m

  7. Service Degradation for GLM 5.3 Flash

    minor ·

    lasted 6m

Fireworks AI website availability

Measured by FlareWarden · fireworks.ai

Last 7 days
100%
Last 30 days
100%
Last 90 days
100%
Avg response, 30 days
568ms

Last 30 days

19-day clean streak
Sep 25: no data
Sep 26: no data
Sep 27: no data
Sep 28: no data
Sep 29: no data
Sep 30: no data
Oct 1: no data
Oct 2: no data
Oct 3: no data
Oct 4: no data
Oct 5: no data
Oct 6: 100.00% uptime, 16 checks
Oct 7: 100.00% uptime, 32 checks
Oct 8: 100.00% uptime, 32 checks
Oct 9: 100.00% uptime, 32 checks
Oct 10: 100.00% uptime, 32 checks
Oct 11: 100.00% uptime, 32 checks
Oct 12: 100.00% uptime, 32 checks
Oct 13: 100.00% uptime, 32 checks
Oct 14: 100.00% uptime, 31 checks
Oct 15: 100.00% uptime, 32 checks
Oct 16: 100.00% uptime, 30 checks
Oct 17: 100.00% uptime, 29 checks
Oct 18: 100.00% uptime, 32 checks
Oct 19: 100.00% uptime, 32 checks
Oct 20: 100.00% uptime, 32 checks
Oct 21: 100.00% uptime, 32 checks
Oct 22: 100.00% uptime, 32 checks
Oct 23: 100.00% uptime, 32 checks
Oct 24: 100.00% uptime, 24 checks
Sep 25 Today
No downtime Partial Downtime Not measurable No data

Measured by FlareWarden's own checks of fireworks.ai from multiple regions every 6 hours. Fireworks AI's status page sets the verdict above when it reports trouble, but never these figures. Percentages count only checks that returned an availability answer — 19 days measured so far. A dash means that window does not yet hold enough measured days to publish a figure.

What Fireworks AI does

Fireworks AI runs open models such as Kimi, DeepSeek, GLM, Qwen and gpt-oss for developers who call them from their own apps and agents. Serverless inference is billed per token on shared Fireworks-managed infrastructure, while on-demand deployments put a model on dedicated GPUs. Teams also train and fine-tune models there. Requests go to api.fireworks.ai, which works with the OpenAI client library.

What an outage looks like

Most incidents hit a single model rather than the whole platform: calls to one serverless model fail or slow down while other models answer normally. Fireworks documents 503 Service Unavailable, 502 and 504 as outage or overload errors. A 429 Too Many Requests usually means your account hit its rate limit, and the Fireworks console may show no traffic for requests that never reached the server.

What to do about it

Open status.fireworks.ai and find the row for the exact model you call, since it lists each serverless model separately. Retry 429, 500, 502, 503 and 504 responses with exponential backoff and jitter, and do not retry 400, 401, 404 or 422 errors. If one model stays degraded, switch the model ID to another serverless model. Report persistent 500s to support@fireworks.ai or the Fireworks Discord.

Is it down for everyone, or just you?

If this page says Fireworks AI is up but it is not loading for you, the problem is between you and them. Run a check against any URL from up to 17 cities across three regions to find out where it breaks.

Test it yourself

Related services

Together AI
Together AI
Groq
Groq
Baseten
Baseten
O
OpenRouter

Fireworks AI outage FAQ

Is Fireworks AI down for everyone or just me?
Check the row for your model on status.fireworks.ai, because Fireworks lists each serverless model separately and most incidents name one model. If that row is green, look at the error code: a 429 is your own account's rate limit and a 401 or 404 is a key or model ID problem. The checks on this page probe fireworks.ai, the company website, so an API fault on api.fireworks.ai can happen while they stay green.
Does the Fireworks status page cover dedicated deployments?
No, it shows uptime only for Serverless, with one row per serverless model. On-demand deployments run on dedicated GPUs, so a 429 there means the deployment's capacity is saturated rather than a platform outage, and Fireworks suggests reducing burst concurrency or adding replicas. For dedicated deployments, Fireworks offers per-deployment Prometheus metrics that show what its infrastructure observed.
What is the difference between a 429 and a 503 from Fireworks?
A 429 means your account went over its serverless rate limit, which Fireworks scopes per account and per model and adapts to your recent usage. A 503 Service Overloaded means the model was busy and shed your request even though you were within your limit. Retry both with exponential backoff. Fireworks says Priority tier requests are prioritized during high load, which lowers the chance of 503s.
Can I switch to another provider during a Fireworks outage?
Yes, if your code uses the OpenAI client library. Fireworks documents connecting the OpenAI client by setting base_url to https://api.fireworks.ai/inference/v1, so moving to another OpenAI-compatible provider means changing that base URL, the API key and the model name. Within Fireworks, switching the model ID to a different serverless model is the quicker fallback when only one model is degraded.
How do I contact Fireworks support during an outage?
Email support@fireworks.ai or post in the Fireworks Discord, both listed in its error code documentation. Fireworks says a 500 Internal Server Error usually needs its engineering team and should be reported right away. Include the full error response and request ID. For more serverless capacity or higher rate limits, Fireworks points to inquiries@fireworks.ai instead.

How we measure this

  • We request Fireworks AI's public endpoint every 6 hours from Fly.io regions across six continents — the most recent check ran from 8 of them. When you open this page and our last check is more than 30 minutes old, we check again on the spot; those checks update the verdict, never the availability figures.
  • A region counts as down only when it gets no usable HTTP response. A 403 or 429 means the origin answered and refused us, which we report as blocked, never as an outage.
  • A single failing region is treated as probe noise. Our checks only change their verdict when two consecutive cycles agree.
  • Response times average only the regions that actually served the page, so a timeout never inflates the number.
  • Where Fireworks AI publishes an official status feed we read it too, and the verdict is the worse of what it reports and what our checks see: a major outage of Fireworks AI itself is Down, trouble with one part of it is Degraded, problems at a few of a provider's many locations change nothing, and a security notice that marks nothing down is shown but changes nothing either. We check the feed every 3 minutes and read it again within minutes of any change to its incidents or components; a reading counts for 45 minutes after the last check that found nothing changed. The availability figures above are our measurement alone.

Get alerted when Fireworks AI goes down

We check Fireworks AI every 6 hours, and again when you open this page if that was more than 30 minutes ago. Your own monitors run as often as every 30 seconds, from up to 17 cities across three regions, and tell you the moment something breaks.

Start Free Monitoring
Free plan available No credit card required