Mixtral-8x22B-Instruct-v0.1: benchmark scores, price and ranking
Mixtral-8x22B-Instruct-v0.1 from Mistral AI ranks #673 of 743 LLMs on the LLMs Tiger aggregate benchmark score (43.5), based on 3 benchmark results captured up to 2026-10-13.
- Organisation
- Mistral AI
- Type
- Open-weight / local model
- Open weights
- Yes
- Released
- 2024-04-17 (source: epoch)
- Cheapest API price
- $0.600 in / $0.600 out per million tokens (via nscale, observed 2026-10-13T20:47Z)
- Aggregate rank
- #673 of 743 (score 43.5)
Relative to other models, Mixtral-8x22B-Instruct-v0.1 is strongest on DTBench (#113 of 146) and weakest on WeirdML (#108 of 109).
Across the 3 benchmarks where at least ten models are scored, Mixtral-8x22B-Instruct-v0.1 is in the top 10% on 0 and in the top half on 0.
Its cheapest listed input price, $0.600 per million tokens, is lower than that of 45% of the 427 models with a listed price.
Mixtral-8x22B-Instruct-v0.1 benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| DTBench | 55.1% | #113 of 146 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Cybench | 7.5% | #15 of 16 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WeirdML | 3.2% | #108 of 109 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Mixtral-8x22B-Instruct-v0.1
Mixtral-8x22B-Instruct-v0.1 is an open-weight model that can run on your own hardware from Mistral AI. It was released on 2024-04-17, according to Epoch AI. The cheapest API listing we track charges $0.600 per million input tokens and $0.600 per million output tokens, through nscale. We hold 3 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where Mixtral-8x22B-Instruct-v0.1 stands
On our aggregate score, Mixtral-8x22B-Instruct-v0.1 is #673 of 743 ranked models, in the lower half. Its score is 43.5; the scale is centred on 50, so it sits 6.5 points below the typical model. That beats 9% of the ranked field. Among the 44 ranked Mistral AI models it is #34. The next one up is mistral-medium-2508 at 43.7. The next one down is Mistral Medium 3.5 at 43.5.
Mixtral-8x22B-Instruct-v0.1, benchmark by benchmark
Mixtral-8x22B-Instruct-v0.1 lands below the middle on DTBench: 55.1%, #113 of 146, with 77% of the field scoring higher. Claude Opus 5.5 (max) leads at 98.9%; the closest result above is mistral-large-2402 at 55.9%. It covers dynamic tool-use benchmark. The board median is 68.8%, so Mixtral-8x22B-Instruct-v0.1 is 13.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Cybench is a weak spot. Mixtral-8x22B-Instruct-v0.1 scores 7.5%, #15 of 16, behind 88% of the models measured. The leader, grok-4-0709, reaches 43.0%. It covers cybersecurity capture-the-flag tasks. The board median is 17.5%, so Mixtral-8x22B-Instruct-v0.1 is 10.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
WeirdML is a weak spot. Mixtral-8x22B-Instruct-v0.1 scores 3.2%, #108 of 109, behind 98% of the models measured. The leader, GPT-6 Astra (max), reaches 93.3%. It covers unusual machine-learning tasks. The board median is 41.2%, so Mixtral-8x22B-Instruct-v0.1 is 38.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
General knowledge and chat: weak, ahead of 4% of the field on average across 2 benchmarks. Its best board here is Cybench (#15 of 16) and its weakest is WeirdML (#108 of 109).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Mixtral-8x22B-Instruct-v0.1 costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on gemma-2-27b ($0.650) and gemma-2-27b-it ($0.800).
Release history
The Mistral AI model released before it was mistral-small-2402 on 2024-02-26, which scores 46.7. The next one was open-mixtral-8x22b on 2024-04-17, at 46.0.
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
- #670 deepseek-r1-0528-qwen3-8b (43.6)
- #671 ollama/granite3-dense:8b (43.6)
- #672 llemma_7b (43.6)
- #674 Mistral Medium 3.5 (43.5)
- #675 Qwen2-1.5B-Instruct (43.4)
- #676 Qwen2-1.5B (43.4)
More from Mistral AI
- Mistral Large 2 (2407) (#231, 49.9)
- codestral-2405 (#352, 48.4)
- pixtral-large-2411 (#360, 48.3)
- Codestral-22B-v0.1-Q4_K_M (#392, 48.0)
- codestral:22b-v0.1-q8_0 (#393, 48.0)
- Mistral Large 4 (high) (#411, 47.7)
Compare it on the live leaderboard → · How the aggregate score works