fix: add missing fireworks models in cost map - #30580
Conversation
…ral voxtral-small-2507 to cost map
The Fireworks model page publishes only a ~1M context length and no separate generation cap, so the entry set max_output_tokens and max_tokens to the full 1048576 window. That equals max_input_tokens, which makes get_modified_max_tokens treat it as a shared input+output budget and lets callers request output sizes the model cannot actually produce. Cap output at 384000 to match the azure_ai/deepseek-v4-pro entry for the same underlying model (identical input/output costs), keeping the Fireworks 1048576 input context. Mirrored in the backup map.
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
|
🚅 Hi, thanks for the PR! I'm Agent Shin, the automated triage bot for this repository. What's this and why am I getting it? I read the description against our contribution rubric. Here's how it lined up: What you got right:
What's still missing:
If the description isn't updated in the next 24 hours, I'll auto-close this PR. That's not us saying we don't care about the change; we want the open-PR list to mirror what a maintainer can act on right now, so contributors don't get lost in a backlog. A closed PR is a soft "park this for later," not a rejection. Take your time; everything below still works after the close. During the grace period: just update the PR description with the missing pieces. No need to ping me; I'll re-check on the next sweep and skip the auto-close if it now passes. See what counts as QA proof for the full rubric (a linked issue alone isn't enough; it covers context, not proof). If the PR does get auto-closed in 24 hours, you still have easy recovery paths:
Internal BerriAI contributors: this rubric doesn't apply to you; ping a maintainer. (I'm an LLM, so I'm not infallible. If you think I got this wrong, ping a maintainer; they'll override me.) |
|
Superseded by rolling registry audit #37902, which reverified and absorbed the provider-documented metadata changes at field level |
2 participants
Copy of #30564 on an in-repo branch so CircleCI runs against it (the original PR's head lives on a fork).
Relevant issues
LIT-3783
added pricing for missing models in cost map for accurate spend tracking
source:
Linear ticket
LIT-3783
Pre-Submission checklist
Please complete all items before asking a LiteLLM maintainer to review your PR
make test-unit@greptileaiand received a Confidence Score of at least 4/5 before requesting a maintainer reviewType
🐛 Bug Fix
Changes
Data-only additions to the model cost map JSON files; no runtime logic changes.
Fireworks AI: new entries for DeepSeek V4 Pro, Kimi K2.6, and Qwen 3.6 Plus, each under both the full
accounts/fireworks/models/...path and the shorterfireworks_ai/...alias. Entries include per-token input/output (and cache-read where applicable), context limits, and capability flags such as reasoning, tools, and vision where relevant. DeepSeek V4 Pro capsmax_output_tokensat 384000 to match theazure_ai/deepseek-v4-proentry for the same underlying model, keeping the Fireworks 1048576 input context.Mistral: adds
mistral/voxtral-small-2507with text pricing, 32K context, andsupports_audio_inputalongside standard chat/tool capabilities.Generated by Claude Code