Gemini 4 Argon
Claude Opus 5.5
Muse Wonder
MiMo-V2.6-Pro
Qwen3.8-Max
Kimi K3
GLM-5.3
GPT-5.5
ERNIE 5.1
DeepSeek-V4.1-Flash-20260910
amazon-nova-experimental-chat
Dola Seed 2.0 Pro
NVIDIA-Nemotron-3-Ultra-550B-A55B
Gemma 4 31B
Step 5 Preview
Grok 4.20
Grok 4.20 is a proprietary model from xAI, released Mar 31, 2026. Its best result is #6 of 96 on AIME with 96.46%. It costs $1.25 per million input tokens and $2.50 per million output tokens, with a 2M-token context window.
- Developer
- xAI
- Released
- Mar 31, 2026
- Context window
- 2M
- Price in / out, per 1M tokens
- $1.25 / $2.50
- Weights
- Proprietary
- Inputs
- Text, Images, Files
- Reasoning
- Yes
- Tool use
- Yes
Benchmarks
| Benchmark | Score bar | Accuracy | Rank | Cost / test |
|---|---|---|---|---|
| SWE-benchCoding | 72.20% | 53 / 85 | $0.52 | |
| Vibe Code BenchCoding | 4.06% | 103 / 115 | $0.77 | |
| LiveCodeBenchCoding | 84.27% | 40 / 143 | $2 / $6 | |
| Tax Agent BenchFinance | 6.95% | 68 / 76 | $0.12 | |
| Legal Research BenchLaw | 13.94% | 73 / 85 | $0.26 | |
| MedCodeMedicine | 32.16% | 91 / 107 | $2 / $6 | |
| MedQAMedicine | 94.55% | 16 / 92 | $2 / $6 | |
| GPQA DiamondReasoning | 88.64% | 34 / 138 | $2 / $6 | |
| MMLU-ProReasoning | 86.25% | 44 / 138 | $2 / $6 | |
| AIMEReasoning | 96.46% | 6 / 96 | $2 / $6 | |
| MMMUReasoning | 83.47% | 32 / 93 | $2 / $6 |
Vals AI, as of Oct 24, 2026.
Compare Grok 4.20
xAI
The lab behind Grok, trained on one of the largest compute clusters and integrated with X. Founded 2023, based in San Francisco Bay Area. x.ai
Questions
Who makes Grok 4.20?
Grok 4.20 is made by xAI, founded 2023 and based in San Francisco Bay Area.
How much does Grok 4.20 cost?
$1.25 per million input tokens and $2.50 per million output tokens, at its provider’s list price.
What is Grok 4.20’s context window?
2M tokens (2,000,000).
Is Grok 4.20 open source?
No. Grok 4.20 is proprietary and served through its provider’s API.