Models, measured side by side
gpt-oss-20b vs Mistral Nemo
Compare gpt-oss-20b, Mistral Nemo by token pricing, context length, benchmark results, speed and tool support.
Monthly workload
Monthly cost breakdown
InputOutput
gpt-oss-20b
$0.81 / mo
Mistral Nemo
$0.53 / mo
Performance comparison
gpt-oss-20b
9
Mistral Nemo
Not listed
Overview
| Overview | gpt-oss-20b | Mistral Nemo |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Released | 2025-08-05 | 2024-07-19 |
| Status | Active | Active |
| Input formats | text | text |
| Output formats | text | text |
Provider
- gpt-oss-20b
- OpenAI
- Mistral Nemo
- Mistral AI
Released
- gpt-oss-20b
- 2025-08-05
- Mistral Nemo
- 2024-07-19
Status
- gpt-oss-20b
- Active
- Mistral Nemo
- Active
Input formats
- gpt-oss-20b
- text
- Mistral Nemo
- text
Output formats
- gpt-oss-20b
- text
- Mistral Nemo
- text
Pricing & capacity
| Pricing & capacity | gpt-oss-20b | Mistral Nemo |
|---|---|---|
| Input / 1M tokens | $0.02 | $0.02 |
| Output / 1M tokens | $0.09 | $0.03 |
| Cached input / 1M | $0.0090 | Not listed |
| Monthly cost | $0.81 | $0.53 |
| Context window | 131K | 131K |
| Maximum output | 33K | 16K |
Input / 1M tokens
- gpt-oss-20b
- $0.02
- Mistral Nemo
- $0.02
Output / 1M tokens
- gpt-oss-20b
- $0.09
- Mistral Nemo
- $0.03
Cached input / 1M
- gpt-oss-20b
- $0.0090
- Mistral Nemo
- Not listed
Monthly cost
- gpt-oss-20b
- $0.81
- Mistral Nemo
- $0.53
Context window
- gpt-oss-20b
- 131K
- Mistral Nemo
- 131K
Maximum output
- gpt-oss-20b
- 33K
- Mistral Nemo
- 16K
Benchmarks & performance
| Benchmarks & performance | gpt-oss-20b | Mistral Nemo |
|---|---|---|
| Intelligence Index | 9 | Not listed |
| GPQA | 68.8% | Not listed |
| Humanity’s Last Exam | 11% | Not listed |
| SciCode | 38.9% | Not listed |
| Output speed | 207 tokens/s | Not listed |
| Time to first token | 0.85 s | Not listed |
Intelligence Index
- gpt-oss-20b
- 9
- Mistral Nemo
- Not listed
GPQA
- gpt-oss-20b
- 68.8%
- Mistral Nemo
- Not listed
Humanity’s Last Exam
- gpt-oss-20b
- 11%
- Mistral Nemo
- Not listed
SciCode
- gpt-oss-20b
- 38.9%
- Mistral Nemo
- Not listed
Output speed
- gpt-oss-20b
- 207 tokens/s
- Mistral Nemo
- Not listed
Time to first token
- gpt-oss-20b
- 0.85 s
- Mistral Nemo
- Not listed
Tools & features
| Tools & features | gpt-oss-20b | Mistral Nemo |
|---|---|---|
| Function calling | Supported | Supported |
| Structured outputs | Supported | Supported |
| JSON mode | Supported | Supported |
| Reasoning | Supported | Not listed |
| Log probabilities | Supported | Supported |
| Deterministic seed | Supported | Supported |
| Prompt caching | Supported | Not listed |
Function calling
- gpt-oss-20b
- Supported
- Mistral Nemo
- Supported
Structured outputs
- gpt-oss-20b
- Supported
- Mistral Nemo
- Supported
JSON mode
- gpt-oss-20b
- Supported
- Mistral Nemo
- Supported
Reasoning
- gpt-oss-20b
- Supported
- Mistral Nemo
- Not listed
Log probabilities
- gpt-oss-20b
- Supported
- Mistral Nemo
- Supported
Deterministic seed
- gpt-oss-20b
- Supported
- Mistral Nemo
- Supported
Prompt caching
- gpt-oss-20b
- Supported
- Mistral Nemo
- Not listed
API & availability
| API & availability | gpt-oss-20b | Mistral Nemo |
|---|---|---|
| API identifier | gpt-oss-20b | mistral-nemo |
| Knowledge cutoff | 2024-06-30 | 2024-04-30 |
| Prices checked | Oct 8, 2026 | Oct 8, 2026 |
API identifier
- gpt-oss-20b
gpt-oss-20b- Mistral Nemo
mistral-nemo
Knowledge cutoff
- gpt-oss-20b
- 2024-06-30
- Mistral Nemo
- 2024-04-30
Prices checked
- gpt-oss-20b
- Oct 8, 2026
- Mistral Nemo
- Oct 8, 2026
About the models
gpt-oss-20b
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...
Full pricing & detailsMistral Nemo
A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA. The model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese,...
Full pricing & detailsComparison FAQ
gpt-oss-20b: $0.02 input and $0.09 output per million tokens. Mistral Nemo: $0.02 input and $0.03 output per million tokens. The cheaper choice depends on your input-to-output ratio; a model can have a lower input rate but a higher output rate.