Kimi K2.5 (Fireworks): benchmark scores, price and ranking
Kimi K2.5 (Fireworks) from Moonshot ranks #50 of 743 LLMs on the LLMs Tiger aggregate benchmark score (52.8), based on 3 benchmark results captured up to 2026-10-13.
- Organisation
- Moonshot
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2026-01-27 (source: epoch)
- Cheapest API price
- $0.450 in / $2.25 out per million tokens (via deepinfra, observed 2026-10-13T20:47Z)
- Aggregate rank
- #50 of 743 (score 52.8)
Relative to other models, Kimi K2.5 (Fireworks) is strongest on OTIS Mock AIME (#36 of 189) and weakest on Chess Puzzles (#63 of 122).
Across the 3 benchmarks where at least ten models are scored, Kimi K2.5 (Fireworks) is in the top 10% on 0 and in the top half on 2.
Its cheapest listed input price, $0.450 per million tokens, is lower than that of 52% of the 427 models with a listed price.
Kimi K2.5 (Fireworks) benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| OTIS Mock AIME | 92.2% | #36 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GPQA Diamond | 87.6% | #57 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 12.0% | #63 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Kimi K2.5 (Fireworks)
Kimi K2.5 (Fireworks) is an open-weight model that can run on your own hardware from Moonshot. It was released on 2026-01-27, according to Epoch AI. The cheapest API listing we track charges $0.450 per million input tokens and $2.25 per million output tokens, through deepinfra. We hold 3 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where Kimi K2.5 (Fireworks) stands
On our aggregate score, Kimi K2.5 (Fireworks) is #50 of 743 ranked models, in the top fifty. Its score is 52.8; the scale is centred on 50, so it sits 2.8 points above the typical model. That beats 93% of the ranked field. Among the 13 ranked Moonshot models it is #2. The next one up is Kimi K3 (max) at 55.7. The next one down is Kimi K2.5 at 50.7.
Kimi K2.5 (Fireworks), benchmark by benchmark
OTIS Mock AIME puts Kimi K2.5 (Fireworks) at 92.2%, #36 of 189. That beats 81% of the models tested, but 7.8 points short of Claude Fable 5.1 (max) (100.0%). The results either side are grok-4.20-0309-reasoning (92.2%) and Gemini 3 Pro Preview (91.4%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Kimi K2.5 (Fireworks) is 34.4 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On GPQA Diamond, Kimi K2.5 (Fireworks) reaches 87.6%. It ranks #57 of 275, ahead of 79% of the field, and 8.2 points behind GPT-6 Astra (max) at 95.8%. The results either side are Kimi K2.5 (87.6%) and Qwen 3.6 Max (Preview) (87.4%). It covers expert-level science questions. The board median is 71.7%, so Kimi K2.5 (Fireworks) is 15.9 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Kimi K2.5 (Fireworks) lands below the middle on Chess Puzzles: 12.0%, #63 of 122, with 51% of the field scoring higher. GPT-6 Astra (max) leads at 72.0%; the closest result above is GPT-5.5 Instant at 12.0%. It covers chess tactics puzzles. The board median is 12.5%, so Kimi K2.5 (Fireworks) is 0.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: strong, ahead of 80% of the field on average across 2 benchmarks. Its best board here is OTIS Mock AIME (#36 of 189) and its weakest is GPQA Diamond (#57 of 275).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Kimi K2.5 (Fireworks) costs less than the typical listing. Models that score within 1.5 points or better and cost less: DeepSeek V4 Flash 0731 (max) ($0.016, 53.1), GLM-5.2 (high) ($0.060, 51.6), Qwen3-Coder-30B-A3B-Instruct ($0.070, 52.7). Similar scores cost more on Gemini 3 Flash ($0.500) and Kimi K2 ($0.550).
Release history
The Moonshot model released before it was Kimi K2.5 on 2026-01-27, which scores 50.7. The next one was Kimi K2.5 (instant) on 2026-01-27, at 49.1.
Reading these numbers
The 3 scores on this page come from 1 source: Epoch AI Benchmarking Hub (3). No official leaderboard supplied any of them, so treat them as third-party measurements. With 3 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #47 Claude Mythos Preview (Early) (53.0)
- #48 Gemini 2.5 Pro Preview 05-06 (53.0)
- #49 Solar-Open2-250B (52.9)
- #51 CodeAct v2.1 (claude-3-5-sonnet-20241022) (52.8)
- #52 claude-3.5-sonnet-20240620 (52.7)
- #53 Qwen3-Coder-30B-A3B-Instruct (52.7)
More from Moonshot
- Kimi K3 (max) (#15, 55.7)
- Kimi K2.5 (#173, 50.7)
- Kimi K2 Thinking Turbo (#187, 50.5)
- Kimi K2 Thinking (Together) (#240, 49.7)
- Kimi-K2-Instruct-0905 (#258, 49.5)
- Kimi K2.5 (instant) (#293, 49.1)
Compare it on the live leaderboard → · How the aggregate score works