Fireworks AI Release notes and new features

Fireworks AI releases Training Cookbook on PyPI and migrates repository

Fireworks AI published fireworks-training-cookbook 2026.11.5.1 on PyPI for Python 3.11+ and migrated its training cookbook to fw-ai-external/cookbook.

Errors identify their source

Fireworks AI updated its error reporting to provide more granular source identification for failures, including new error codes for parsing and transport issues.

Errors identify their source

Before 1.2.27, the sidecar returned HTTP 502 for two failures that weren't server-side: tito_model_malformed_output (unparseable model output) and tito_upstream_error (any sampling exception, including client timeouts and dropped connections). They are replaced by: tito_sdk_parse_error (422): the output couldn't be parsed into an assistant or tool message.

Action required

If you use any of the models below on serverless, migrate to a recommended replacement before November 7, 2026 . After that date, they will no longer be available via serverless endpoints.

• Action required
• New models on Serverless
• Recommended migrations

Inference Doubled serverless rate limits for Small models We doubled the adaptive rate-limit ceilings for...

Inference Doubled serverless rate limits for Small models We doubled the adaptive rate-limit ceilings for Small serverless models (less than 600B total parameters). The Small tier now has these ceilings: Total Prompt TPM: 216M Uncached Prompt TPM: 54M Generated TPM: 2.16M Medium and Large model ceilings are unchanged.

Inference Higher serverless rate limits for Small models We increased the adaptive rate-limit ceilings for...

Inference Higher serverless rate limits for Small models We increased the adaptive rate-limit ceilings for Small serverless models and expanded the Small tier to cover models with less than 600B total parameters .

Training New Serverless Training model: GLM 5.3 Flash GLM 5.3 Flash is now available for LoRA workloads on...

Training New Serverless Training model: GLM 5.3 Flash GLM 5.3 Flash is now available for LoRA workloads on the shared Serverless Training pool, with up to 200K context and both text and vision inputs. See the Serverless Training guide for setup and the training model catalog for current availability.

Recommended migrations

- DeepSeek V4 Pro (0813) , migrate to DeepSeek V4.1 Flash DeepSeek V4 Flash (0731) , migrate to DeepSeek V4.1 Flash DeepSeek V4 Flash Vision Exp , migrate to DeepSeek V4.1 Flash Muse Glimmer 30B , migrate to NVIDIA Nemotron 3.5 Lightning 30B A3B Kimi K2.6 , migrate to GLM 5.3 or Kimi K3 Kimi K2.7 Code , migrate to GLM 5.3 or Kimi K3 GLM 5.2 , including GLM...

Action required

If you use any of the models below on serverless, migrate to a recommended replacement before September 25, 2026 . After that date, they will no longer be available via serverless endpoints.

• Action required
• Recommended migrations

Training Training cost estimator The new training cost estimator helps you estimate what a training job will...

Training Training cost estimator The new training cost estimator helps you estimate what a training job will cost before you run it. Managed and Serverless estimates use published per-token rates. Dedicated estimates use allocated GPU-hour rates. Planning estimates are not quotes.

Platform Deployment tags and annotation API changes Deployment tags are customer-managed entries stored in a...

Platform Deployment tags and annotation API changes Deployment tags are customer-managed entries stored in a deployment's annotations map. firectl presents logical keys such as environment ; the REST API represents the same key as custom/environment . firectl: Version 1.8.3 adds deployment tag set , unset , and list , including atomic batch operations.

Inference Serverless rate limit ceilings now scale by model size Serverless adaptive rate limit ceilings now...

Inference Serverless rate limit ceilings now scale by model size Serverless adaptive rate limit ceilings now vary by model size tier. Smaller models ( Serverless rate limits for tier thresholds and ceiling values.

Training New Serverless Training models: DeepSeek V4 Flash 0731, Qwen 3.8 27B, and Muse Glimmer 30B The...

Training New Serverless Training models: DeepSeek V4 Flash 0731, Qwen 3.8 27B, and Muse Glimmer 30B The following models are now available for LoRA workloads on the shared Serverless Training pool: DeepSeek V4 Flash 0731 with up to 262K context Qwen 3.8 27B with up to 128K context Muse Glimmer 30B with up to 128K context See the Serverless Training guide...

Recommended migrations

- MiniMax M2.7 , migrate to MiniMax M3 GPT OSS 20B , migrate to GPT OSS 120B or Qwen3 8B for lower-latency workloads Kimi K2.6 Turbo / Fast , migrate to Kimi K2.6 (standard mode) Kimi K2.7 Code Fast , migrate to Kimi K2.7 Code (standard mode) DeepSeek V4 Pro , migrate to DeepSeek V4 Pro (0813)

Platform SSO documentation: IdP-initiated SAML Updated the Custom SSO guide

Platform SSO documentation: IdP-initiated SAML Updated the Custom SSO guide. IdP-initiated SAML is supported as an opt-in ( --enable-idp-initiated-sso ); the previous troubleshooting copy that said Fireworks only supported SP-initiated login was incorrect.

Training SFT learning rate scheduler documentation Documented learning rate scheduler settings for supervised...

Training SFT learning rate scheduler documentation Documented learning rate scheduler settings for supervised fine-tuning jobs, including constant , linear , and cosine schedules via firectl and the REST API lrScheduler object.