Mixtral 8x22B Instruct vs Mistral Large

Mistral Large is the better all-round choice, scoring 44/100 against Mixtral 8x22B Instruct (34). It leads on price, inputs & features and context window. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Mistral Large is cheaper at $0.50 input / $1.50 output per million tokens (official Mistral API price). Mixtral 8x22B Instruct costs $2.00 input / $6.00 output per million tokens (median across 5 API providers). At a typical mix of three input tokens to one output token, that is $0.75 per million tokens for Mistral Large versus $3.00 for Mixtral 8x22B Instruct (4× as much).

Mistral Large scores higher on the Capabilities Index (ECI): Mistral Large 122.0 (#153 of 165) and Mixtral 8x22B Instruct 122.0 (#155 of 165). The confidence ranges of the top two overlap (115.7–124.8 vs 115.7–124.3), so treat the gap as small. On individual benchmarks: GPQA Diamond — Mistral Large 38.8%, Mixtral 8x22B Instruct 34.1%.

There are no published SWE-bench Verified results for Mixtral 8x22B Instruct and Mistral Large yet, so there is no like-for-like coding score. On overall capability, Mistral Large leads, which tends to carry over to coding, but test on your own codebase. Both support tool calling for agent workflows.

Mistral Large has the largest context window at 128,000 tokens, against 65,536 for Mixtral 8x22B Instruct. Maximum output per response: Mixtral 8x22B Instruct up to 52,428, Mistral Large up to 102,400 tokens.

Mixtral 8x22B Instruct accepts text and PDFs; Mistral Large accepts text and PDFs. They handle the same number of input types.

Mixtral 8x22B Instruct publishes its weights and can be self-hosted; Mistral Large is proprietary.

Mixtral 8x22B Instruct is the newest, released Apr 17, 2024. Mistral Large came out Feb 26, 2024. Knowledge cutoff: Mixtral 8x22B Instruct Jan 31, 2024, Mistral Large Nov 30, 2024.

Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 2, 2026.