feat: add Cerebras provider integration #6842
Open
+230
−0
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add Cerebras as a new AI provider with: - OpenAI-compatible API endpoint (https://api.cerebras.ai/v1) - 8 models including Llama 4 Scout, Llama 3.3 70B, GPT OSS 120B, and Qwen variants - Tool calling support for gpt-oss-120b and llama-3.3-70b - Ultra-fast inference speeds (2000-3000 tokens/s) - Complete documentation with setup guide and troubleshooting
All 8 Cerebras models support tool calling according to their official documentation. Updated capabilities to include 'tools' for: - llama-4-scout-17b-16e-instruct - llama3.1-8b - qwen-3-32b - qwen-3-235b-a22b-instruct-2507 - qwen-3-235b-a22b-thinking-2507 - qwen-3-coder-480b Also corrected Llama 4 Scout parameter count from 109B to 17B.
…n all models - Corrected Llama 4 Scout parameter count from 109B to 17B - Added tool calling support notation for all 8 models - Updated Features section to list all models with tool calling capability
Disable tool calling for 5 Cerebras models that reject JSON schema
validation fields (minimum, maximum, default). Only gpt-oss-120b,
llama-3.3-70b, and qwen-3-coder-480b support tools with Jan's RAG
tool schemas.
Root cause: Models have inconsistent JSON schema validation strictness.
Most models reject requests containing unsupported fields like minimum/maximum
in tool parameter schemas, while 3 models are more lenient.
Error returned by strict models:
"Unsupported JSON schema fields: {'maximum', 'minimum'}"
Models with tools disabled:
- llama-4-scout-17b-16e-instruct
- llama3.1-8b
- qwen-3-32b
- qwen-3-235b-a22b-instruct-2507
- qwen-3-235b-a22b-thinking-2507
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What
Adds Cerebras as a new AI provider to Jan with support for 8 models, including Llama 4 Scout, Llama 3.3 70B, GPT OSS 120B, and Qwen variants.
Why
Cerebras offers ultra-fast AI inference (2000-3000 tokens/s) with OpenAI-compatible endpoints, making it an excellent addition to Jan's provider ecosystem. This integration enables users to leverage Cerebras's high-performance models directly through Jan's interface.
How
Provider Configuration:
predefinedProvidersarray inweb-app/src/consts/providers.tshttps://api.cerebras.ai/v1Visual Assets:
getProviderLogo()function inweb-app/src/lib/utils.tsDocumentation:
docs/src/pages/docs/desktop/remote-models/cerebras.mdx_meta.jsonTechnical Approach:
This is a configuration-driven integration that leverages Jan's existing OpenAI-compatible provider infrastructure. No custom code or API handlers needed - everything works through the standard token.js fallback mechanism.
Testing
Manual Testing Required:
Automated Testing:
TypeScript compilation verified - no errors in modified files.
Breaking Changes
None. This is a purely additive change that doesn't modify existing provider behavior.
Files Changed
web-app/src/consts/providers.ts- Provider configuration (+91 lines)web-app/src/lib/utils.ts- Logo reference (+2 lines)web-app/public/images/model-provider/cerebras.png- Provider logo (53KB PNG)docs/src/pages/docs/desktop/remote-models/cerebras.mdx- Documentation (new file)docs/src/pages/docs/desktop/remote-models/_meta.json- Navigation metadata (+3 lines)Additional Notes
Model performance specs based on Cerebras documentation:
API compatibility: OpenAI-compatible but does not support
frequency_penalty,logit_bias,presence_penalty,parallel_tool_calls, orservice_tier.