token.app › Compare › gpt-oss-20b vs Grok 4.6
gpt-oss-20b vs Grok 4.6
Grok 4.6 leads on 2 of 2 shared benchmarks. gpt-oss-20b is 83× cheaper per token on a blended 3:1 basis. Grok 4.6 accepts the larger context window (500K).
Head to head
Blended price uses a 3:1 input:output mix
| gpt-oss-20b | Grok 4.6 | |
|---|---|---|
| Provider | OpenAI | SpaceXAI |
| Input $/1M | $0.018 | $2.00 |
| Output $/1M | $0.090 | $6.00 |
| Blended $/1M | $0.036 | $3.00 |
| Context window | 131K | 500K |
| Max output | 33K | 450K |
| Released | Aug 2025 | Aug 2026 |
| Licence | Open weights | Proprietary |
| GPQA Diamond | 60.8% | 94.0% |
| OTIS Mock AIME | 65.3% | 99.2% |
Benchmark scores independently run and published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0. Only benchmarks with a score for BOTH models are compared.
Choose gpt-oss-20b if…
- Costs less per token — $0.036 vs $3.00 blended
- Open weights — self-hostable
Choose Grok 4.6 if…
- Higher GPQA Diamond (94.0% vs 60.8%)
- Higher OTIS Mock AIME (99.2% vs 65.3%)
- Larger context window (500K)
- Newer model (released Aug 2026)
FAQ
Which is better, gpt-oss-20b or Grok 4.6?
Across the 2 benchmarks both models have independently-run scores for, gpt-oss-20b leads on 0 and Grok 4.6 leads on 2. Benchmarks are one input — price and context window are on this page too.
Is gpt-oss-20b cheaper than Grok 4.6?
gpt-oss-20b is cheaper: $0.036 versus $3.00 per million tokens blended at 3:1 input:output. Input alone: $0.018 vs $2.00. Output alone: $0.090 vs $6.00.
What context windows do gpt-oss-20b and Grok 4.6 have?
gpt-oss-20b: 131,072 tokens. Grok 4.6: 500,000 tokens.
Who makes gpt-oss-20b and Grok 4.6?
gpt-oss-20b is made by OpenAI. Grok 4.6 is made by SpaceXAI.