Firesearch: pricing, context window and benchmarks
Firesearch is a mid-range model from Fireworks AI. It offers a 32,768-token context window, costs $7.03 per million input tokens and $29.81 per million output tokens, and averages 68.5 across the index's benchmark suite.
Pricing
Firesearch costs $7.03 per million input tokens and $29.81 per million output tokens. A blended 3:1 input-to-output mix works out to about $12.72 per million tokens. Cached input is listed at $0.74 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $7.03 | per 1M tokens |
| Output | $29.81 | per 1M tokens |
| Cached input | $0.74 | per 1M tokens |
| Blended 3:1 | $12.72 | estimate per 1M tokens |
| 1M in + 1M out | $36.84 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 71.6 | 384 | 36th | |
| GPQA Diamond | 65.0 | 359 | 40th | |
| SWE-bench Verified | 55.6 | 380 | 37th | |
| HumanEval | 73.6 | 365 | 39th | |
| MATH-500 | 68.4 | 394 | 34th | |
| AIME 2025 | 59.1 | 393 | 35th | |
| LiveCodeBench | 57.1 | 393 | 35th | |
| IFEval | 78.1 | 367 | 39th | |
| TruthfulQA | 75.1 | 328 | 45th | |
| HellaSwag | 82.4 | 338 | 44th | |
| MMMU | 67.4 | 353 | 41th | |
| Arena Elo | 1353.2 | 358 | 40th |
What Firesearch is good for
Firesearch is a general-purpose model in the index. It handles text generation, summarization and question answering, and it can be a sensible default when you need a balance of quality, price and context. At $7.03 per million input tokens with a 32,768-token window, check its benchmark profile against the specific task before committing.
Strengths
- Accepts more than text, including image input where the provider supports it.
Trade-offs
- Proprietary licence: you cannot download the weights or self-host the model.
- Average benchmark score is below the index median; test it on your own task.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Claude Haiku 4.5 Nano | Anthropic | 8192 | $6.47 | 68.7 |
| Firesearch | Fireworks AI | 32768 | $7.03 | 68.5 |
| Molmo | Allen Institute for AI | 32768 | $2.49 | 68.5 |
| Mixtral Groq Base | Groq | 32768 | $2.66 | 68.4 |
| Hunyuan-Turbo Mini | Tencent | 8192 | $6.59 | 68.3 |
| Molmo Instruct | Allen Institute for AI | 32768 | $2.53 | 68.3 |
FAQ
How much does Firesearch cost?
Firesearch is listed at $7.03 per million input tokens and $29.81 per million output tokens. Cached input is $0.74 per million tokens. A blended 3:1 input-to-output mix is about $12.72 per million tokens.
What is the context window of Firesearch?
Firesearch has a context window of 32,768 tokens and can emit up to 32,768 output tokens per request in this index.
Is Firesearch open weights?
No. Firesearch is a proprietary model with a proprietary licence, so you cannot download the weights.
How does Firesearch score on benchmarks?
Firesearch averages 68.5 across the scored benchmarks in this index, ranking #378 of 600 models. Its strongest relative result is Arena Elo at 1353.2.
What is Firesearch best for?
Firesearch is a general-purpose model in the index. It handles text generation, summarization and question answering, and it can be a sensible default when you need a balance of quality, price and context. At $7.03 per million input tokens with a 32,768-token window, check its benchmark profile against the specific task before committing.