Mistral Small 4 vs Grok 4.20 (Non-Reasoning) vs Nemotron 3 Super 120B A12B
Mistral Small 4 comes out ahead, 65 to 54 and 52 on our weighted score, and it is the cheaper option too.
- Our pick
Mistral AIMistral Small 4
65/100- ECI—
- Price$0.15 / $0.60
- Context256K
xAIGrok 4.20 (Non-Reasoning)
52/100- ECI—
- Price$1.25 / $2.50
- Context1M
NVIDIANemotron 3 Super 120B A12B
54/100- ECI—
- Price$0.15 / $0.65
- Context262K
Mistral Small 4 is our pick
Mistral Small 4 is the better all-round choice, scoring 65/100 against Nemotron 3 Super 120B A12B (54) and Grok 4.20 (Non-Reasoning) (52). Grok 4.20 (Non-Reasoning) wins on context window. 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.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceMistral Small 4Mistral Small 4 $0.263 · Nemotron 3 Super 120B A12B $0.275 · Grok 4.20 (Non-Reasoning) $1.56 per 1M tokens (3:1 blend)
- Longest contextGrok 4.20 (Non-Reasoning)Grok 4.20 (Non-Reasoning) 1,000,000 · Nemotron 3 Super 120B A12B 262,144 · Mistral Small 4 256,000 tokens
- Widest inputsGrok 4.20 (Non-Reasoning)Mistral Small 4: Text, Images · Grok 4.20 (Non-Reasoning): Text, Images, PDFs · Nemotron 3 Super 120B A12B: Text
- Self-hostingMistral Small 4 and Nemotron 3 Super 120B A12BPublishes downloadable weights
| Measure | Weight | Mistral Small 4 | Grok 4.20 (Non-Reasoning) | Nemotron 3 Super 120B A12B |
|---|---|---|---|---|
| Price | 50% | 74 | 39 | 73 |
| Inputs & features | 30% | 70 | 70 | 35 |
| Context window | 20% | 36 | 60 | 37 |
| Overall | 100% | 65/100 | 52/100 | 54/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | Mistral Small 4Mistral AI | Grok 4.20 (Non-Reasoning)xAI | Nemotron 3 Super 120B A12BNVIDIA |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.15 (best) | $1.25 | $0.15 (best) |
| Output | $0.60 (best) | $2.50 | $0.65 |
| Cached input | $0.015 (best) | $0.20 | — |
| Blended (3:1) | $0.263 (best) | $1.56 | $0.275 |
| Long-context rate | Same rate | Over 200K: $2.50 / $5.00 | Same rate |
| Price source | Official Mistral API | Official xAI API | Median of 16 providers |
| Limits | |||
| Context window | 256,000 tokens | 1,000,000 tokens (best) | 262,144 tokens |
| Max output | 256,000 tokens | 30,000 tokens | 262,144 tokens (best) |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | No |
| PDFs | No | Yes | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yeshigh | No | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Yes | No |
| Availability | |||
| Weights | Open | Proprietary | Open |
| API model ID | mistral-small-2603 | grok-4.20-0309-non-reasoning | — |
| API providers | 21 (best) | 13 | 18 |
| Released | Mar 16, 2026 | Mar 9, 2026 | Mar 11, 2026 |
| Knowledge cutoff | Jun 2025 | — | Jun 2025 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- Mistral Small 4$2.70
- Grok 4.20 (Non-Reasoning)$17.50
- Nemotron 3 Super 120B A12B$2.80
Which should you choose?
Which is better: Mistral Small 4, Grok 4.20 (Non-Reasoning) or Nemotron 3 Super 120B A12B?
Mistral Small 4 is the better all-round choice, scoring 65/100 against Nemotron 3 Super 120B A12B (54) and Grok 4.20 (Non-Reasoning) (52). Grok 4.20 (Non-Reasoning) wins on context window. 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.
Which is cheaper, Mistral Small 4, Grok 4.20 (Non-Reasoning) or Nemotron 3 Super 120B A12B?
Mistral Small 4 is cheaper at $0.15 input / $0.60 output per million tokens (official Mistral API price). Nemotron 3 Super 120B A12B costs $0.15 input / $0.65 output per million tokens (median across 16 API providers); Grok 4.20 (Non-Reasoning) costs $1.25 input / $2.50 output per million tokens (official xAI API price). At a typical mix of three input tokens to one output token, that is $0.263 per million tokens for Mistral Small 4 versus $0.275 for Nemotron 3 Super 120B A12B (1× as much) and $1.56 for Grok 4.20 (Non-Reasoning) (6× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Mistral Small 4 has not been scored yet, Grok 4.20 (Non-Reasoning) has not been scored yet and Nemotron 3 Super 120B A12B has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Mistral Small 4, Grok 4.20 (Non-Reasoning) and Nemotron 3 Super 120B A12B 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.
Which has the bigger context window?
Grok 4.20 (Non-Reasoning) has the largest context window at 1,000,000 tokens, against 262,144 for Nemotron 3 Super 120B A12B and 256,000 for Mistral Small 4. Maximum output per response: Mistral Small 4 up to 256,000, Grok 4.20 (Non-Reasoning) up to 30,000, Nemotron 3 Super 120B A12B up to 262,144 tokens.
Which can read images, PDFs, audio or video?
Mistral Small 4 accepts text and images; Grok 4.20 (Non-Reasoning) accepts text, images and PDFs; Nemotron 3 Super 120B A12B accepts text. Grok 4.20 (Non-Reasoning) handles the widest range of inputs.
Are any of these open source?
Mistral Small 4 and Nemotron 3 Super 120B A12B publishes its weights and can be self-hosted; Grok 4.20 (Non-Reasoning) is proprietary.
Which is newer?
Mistral Small 4 is the newest, released Mar 16, 2026. Nemotron 3 Super 120B A12B came out Mar 11, 2026; Grok 4.20 (Non-Reasoning) came out Mar 9, 2026. Knowledge cutoff: Mistral Small 4 Jun 2025, Nemotron 3 Super 120B A12B Jun 2025.
How do you decide the winner?
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.
