Gemma 3 27B IT vs Llama 4 Scout 17B Instruct vs Ministral 3 8B

Llama 4 Scout 17B Instruct is the better all-round choice, scoring 72/100 against Ministral 3 8B (68) and Gemma 3 27B IT (62). It leads on context window. Ministral 3 8B wins on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

Ministral 3 8B is cheaper at $0.15 input / $0.15 output per million tokens (official Mistral API price). Gemma 3 27B IT costs $0.11 input / $0.30 output per million tokens (median across 16 API providers); Llama 4 Scout 17B Instruct costs $0.17 input / $0.59 output per million tokens (median across 13 API providers). At a typical mix of three input tokens to one output token, that is $0.15 per million tokens for Ministral 3 8B versus $0.157 for Gemma 3 27B IT (1× as much) and $0.275 for Llama 4 Scout 17B Instruct (1.8× as much).

There is no independent benchmark that covers all three models yet. Gemma 3 27B IT has an ECI of 129.9, Llama 4 Scout 17B Instruct has an ECI of 129.6 and Ministral 3 8B has not been scored yet.

There are no published SWE-bench Verified results for Gemma 3 27B IT, Llama 4 Scout 17B Instruct and Ministral 3 8B yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.

Llama 4 Scout 17B Instruct has the largest context window at 10,000,000 tokens, against 262,144 for Ministral 3 8B and 131,072 for Gemma 3 27B IT. Maximum output per response: Gemma 3 27B IT up to 131,072, Llama 4 Scout 17B Instruct up to 16,384, Ministral 3 8B up to 262,144 tokens.

Gemma 3 27B IT accepts text and images; Llama 4 Scout 17B Instruct accepts text and images; Ministral 3 8B accepts text and images. They handle the same number of input types.

Yes, all three publish their weights (Apache 2.0), so you can self-host them.

Ministral 3 8B is the newest, released Dec 2, 2025. Llama 4 Scout 17B Instruct came out Apr 5, 2025; Gemma 3 27B IT came out Mar 12, 2025. Knowledge cutoff: Gemma 3 27B IT Aug 2024, Llama 4 Scout 17B Instruct Aug 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.