Gemma 3 27B IT vs Llama 4 Scout 17B Instruct vs Ministral 3B
Llama 4 Scout 17B Instruct comes out ahead, 58 to 46 and 26 on our weighted score, though Ministral 3B is 2.8× cheaper per token.
Price in rating: Excluded · Include price
Google
Gemma 3 27B IT
46/100- ECI127.5
- Price$0.10 / $0.299
- Context131K
- Our pick
Meta
Llama 4 Scout 17B Instruct
58/100- ECI127.8
- Price$0.17 / $0.59
- Context10M
Mistral AI
Ministral 3B
26/100- ECI—
- Price$0.10 / $0.10
- Context128K
Llama 4 Scout 17B Instruct is our pick
Llama 4 Scout 17B Instruct is the better all-round choice, scoring 58/100 against Gemma 3 27B IT (46) and Ministral 3B (26). It leads on capability and context window. The score weighs capability 67%, inputs & features 20%, context window 13%. Capability uses the average of shared benchmarks (GPQA Diamond), because Ministral 3B has no Capabilities Index score yet.
- CapabilityLlama 4 Scout 17B InstructShared benchmarks: Llama 4 Scout 17B Instruct 51.8% · Gemma 3 27B IT 47.7% · Ministral 3B 25.3%
- Lowest priceMinistral 3BMinistral 3B $0.10 · Gemma 3 27B IT $0.15 · Llama 4 Scout 17B Instruct $0.275 per 1M tokens (3:1 blend)
- Longest contextLlama 4 Scout 17B InstructLlama 4 Scout 17B Instruct 10,000,000 · Gemma 3 27B IT 131,072 · Ministral 3B 128,000 tokens
- Widest inputsGemma 3 27B IT and Llama 4 Scout 17B InstructGemma 3 27B IT: Text, Images · Llama 4 Scout 17B Instruct: Text, Images · Ministral 3B: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Gemma 3 27B IT | Llama 4 Scout 17B Instruct | Ministral 3B |
|---|---|---|---|---|
| CapabilityShared benchmarks | 67% | 48 | 52 | 25 |
| Inputs & features | 20% | 50 | 50 | 25 |
| Context window | 13% | 32 | 100 | 32 |
| Overall | 100% | 46/100 | 58/100 | 26/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 127.5 | 127.8 (best) | — |
| ECI rank | #141 of 158 | #140 of 158 (best) | — |
| GPQA DiamondGraduate-level science questions | 47.7% | 51.8% (best) | 25.3% |
| OTIS Mock AIME 2024–2025Competition mathematics | 22.5% (best) | 7.8% | — |
| Price per million tokens | |||
| Input | $0.10 (best) | $0.17 | $0.10 (best) |
| Output | $0.299 | $0.59 | $0.10 (best) |
| Cached input | — | — | — |
| Blended (3:1) | $0.15 | $0.275 | $0.10 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 17 providers | Median of 13 providers | Median of 1 providers |
| Limits | |||
| Context window | 131,072 tokens | 10,000,000 tokens (best) | 128,000 tokens |
| Max output | 131,072 tokens (best) | 16,384 tokens | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | No | No | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | — | — | — |
| API providers | 18 (best) | 13 | 1 |
| Released | Mar 12, 2025 | Apr 5, 2025 | Oct 16, 2024 |
| Knowledge cutoff | Aug 2024 | Aug 2024 | Mar 2024 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Gemma 3 27B IT$1.60
Llama 4 Scout 17B Instruct$2.88
Ministral 3B$1.20
Which should you choose?
Which is better: Gemma 3 27B IT, Llama 4 Scout 17B Instruct or Ministral 3B?
Llama 4 Scout 17B Instruct is the better all-round choice, scoring 58/100 against Gemma 3 27B IT (46) and Ministral 3B (26). It leads on capability and context window. The score weighs capability 67%, inputs & features 20%, context window 13%. Capability uses the average of shared benchmarks (GPQA Diamond), because Ministral 3B has no Capabilities Index score yet.
Which is cheaper, Gemma 3 27B IT, Llama 4 Scout 17B Instruct or Ministral 3B?
Ministral 3B is cheaper at $0.10 input / $0.10 output per million tokens (median across 1 API provider). Gemma 3 27B IT costs $0.10 input / $0.299 output per million tokens (median across 17 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 $0.10 per million tokens for Ministral 3B versus $0.15 for Gemma 3 27B IT (1.5× as much) and $0.275 for Llama 4 Scout 17B Instruct (2.8× as much).
Which scores higher on benchmarks?
Not every model here has a Capabilities Index score, so we compare the average of shared benchmarks (GPQA Diamond): Llama 4 Scout 17B Instruct 51.8%, Gemma 3 27B IT 47.7% and Ministral 3B 25.3%. On individual benchmarks: GPQA Diamond — Llama 4 Scout 17B Instruct 51.8%, Gemma 3 27B IT 47.7%, Ministral 3B 25.3%.
Which is better for coding?
There are no published SWE-bench Verified results for Gemma 3 27B IT, Llama 4 Scout 17B Instruct and Ministral 3B yet, so there is no like-for-like coding score. On overall capability, Llama 4 Scout 17B Instruct leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
Llama 4 Scout 17B Instruct has the largest context window at 10,000,000 tokens, against 131,072 for Gemma 3 27B IT and 128,000 for Ministral 3B. Maximum output per response: Gemma 3 27B IT up to 131,072, Llama 4 Scout 17B Instruct up to 16,384, Ministral 3B up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Gemma 3 27B IT accepts text and images; Llama 4 Scout 17B Instruct accepts text and images; Ministral 3B accepts text. Gemma 3 27B IT handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights, so you can self-host them.
Which is newer?
Llama 4 Scout 17B Instruct is the newest, released Apr 5, 2025. Gemma 3 27B IT came out Mar 12, 2025; Ministral 3B came out Oct 16, 2024. Knowledge cutoff: Gemma 3 27B IT Aug 2024, Llama 4 Scout 17B Instruct Aug 2024, Ministral 3B Mar 2024.
How do you decide the winner?
This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, 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 18, 2026.