grok-4.20-0309-reasoning: benchmark scores, price and ranking
grok-4.20-0309-reasoning from xAI ranks #249 of 743 LLMs on the LLMs Tiger aggregate benchmark score (49.6), based on 7 benchmark results captured up to 2026-10-13.
- Organisation
- xAI
- Type
- Frontier model
- Open weights
- No
- Released
- 2026-02-17 (source: epoch)
- Cheapest API price
- $1.25 in / $2.50 out per million tokens (via xai, observed 2026-10-13T20:47Z)
- Aggregate rank
- #249 of 743 (score 49.6)
Relative to other models, grok-4.20-0309-reasoning is strongest on GPQA Diamond (#40 of 275) and weakest on FrontierMath T4 (#38 of 50).
Across the 7 benchmarks where at least ten models are scored, grok-4.20-0309-reasoning is in the top 10% on 0 and in the top half on 3.
Its cheapest listed input price, $1.25 per million tokens, is lower than that of 26% of the 427 models with a listed price.
grok-4.20-0309-reasoning benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| GPQA Diamond | 89.3% | #40 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 92.2% | #35 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 24.0% | #35 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T1-3 | 44.9% | #42 of 68 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ProofBench | 14.0% | #33 of 46 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SimpleQA Verified | 30.2% | #51 of 69 | Epoch AI Benchmarking Hub | 2026-10-13 |
| FrontierMath T4 | 17.1% | #38 of 50 | Epoch AI Benchmarking Hub | 2026-10-13 |
About grok-4.20-0309-reasoning
grok-4.20-0309-reasoning is a frontier model from xAI. It was released on 2026-02-17, according to Epoch AI. The cheapest API listing we track charges $1.25 per million input tokens and $2.50 per million output tokens, through xai. We hold 7 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where grok-4.20-0309-reasoning stands
On our aggregate score, grok-4.20-0309-reasoning is #249 of 743 ranked models, around the middle of the table. Its score is 49.6; the scale is centred on 50, so it sits 0.4 points below the typical model. That beats 66% of the ranked field. Among the 26 ranked xAI models it is #11. The next one up is grok-4-20 at 50.0. The next one down is Grok-2 at 49.6.
grok-4.20-0309-reasoning, benchmark by benchmark
GPQA Diamond puts grok-4.20-0309-reasoning at 89.3%, #40 of 275. That beats 85% of the models tested, but 6.4 points short of GPT-6 Astra (max) (95.8%). The results either side are DeepSeek v4 Pro (max) (89.6%) and Ornith-1.5-35B-A3B (89.2%). It covers expert-level science questions. The board median is 71.7%, so grok-4.20-0309-reasoning is 17.6 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On OTIS Mock AIME, grok-4.20-0309-reasoning reaches 92.2%. It ranks #35 of 189, ahead of 81% of the field, and 7.8 points behind Claude Fable 5.1 (max) at 100.0%. The results either side are Qwen3.7 Plus (93.3%) and Kimi K2.5 (Fireworks) (92.2%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so grok-4.20-0309-reasoning is 34.4 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
grok-4.20-0309-reasoning manages 24.0% on Chess Puzzles, good for #35 of 122. GPT-6 Astra (max) leads at 72.0%, 48.0 points further up, and 71% of models sit below. The results either side are GPT-5.4 mini (xhigh) (24.0%) and Qwen3.7 Plus (24.0%). It covers chess tactics puzzles. The board median is 12.5%, so grok-4.20-0309-reasoning is 11.5 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
grok-4.20-0309-reasoning lands below the middle on FrontierMath T1-3: 44.9%, #42 of 68, with 60% of the field scoring higher. GPT-6 Astra (max) leads at 93.7%; the closest result above is DeepSeek v4 Pro (max) at 45.3%. It covers research-level maths, tiers 1-3. The board median is 55.6%, so grok-4.20-0309-reasoning is 10.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
grok-4.20-0309-reasoning lands below the middle on ProofBench: 14.0%, #33 of 46, with 70% of the field scoring higher. Claude Fable 5.1 (max) leads at 100.0%; the closest result above is GPT-5.2 (xhigh) at 15.0%. It covers writing mathematical proofs. The board median is 21.0%, so grok-4.20-0309-reasoning is 7.0 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
grok-4.20-0309-reasoning lands below the middle on SimpleQA Verified: 30.2%, #51 of 69, with 72% of the field scoring higher. gemini-4-argon-high leads at 76.0%; the closest result above is Claude Sonnet 4.5 (59k thinking) at 30.7%. It covers short factual questions. The board median is 41.0%, so grok-4.20-0309-reasoning is 10.8 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
grok-4.20-0309-reasoning lands below the middle on FrontierMath T4: 17.1%, #38 of 50, with 74% of the field scoring higher. GPT-6.1 Sol (max) leads at 100.0%; the closest result above is Grok 4.7 (xhigh) at 17.1%. It covers research-level maths, tier 4. The board median is 30.5%, so grok-4.20-0309-reasoning is 13.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: solid, ahead of 51% of the field on average across 5 benchmarks. Its best board here is GPQA Diamond (#40 of 275) and its weakest is FrontierMath T4 (#38 of 50). General knowledge and chat: uneven, ahead of 49% of the field on average across 2 benchmarks. Its best board here is Chess Puzzles (#35 of 122) and its weakest is SimpleQA Verified (#51 of 69).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so grok-4.20-0309-reasoning costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), DeepSeek V4 Flash 0731 (max) ($0.016, 53.1). Similar scores cost more on Qwen3.8 Max (0902) (xhigh) ($1.65) and Gemini 3 Pro ($2.00).
Release history
The xAI model released before it was grok-4-20 on 2026-02-17, which scores 50.0. The next one was Grok 4.20 (Reasoning) on 2026-02-17, at 51.2.
Reading these numbers
The 7 scores on this page come from 1 source: Epoch AI Benchmarking Hub (7). No official leaderboard supplied any of them, so treat them as third-party measurements. With 7 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #246 GLM-5.1 (49.6)
- #247 Opus 4.5 (Thinking, 64K) (49.6)
- #248 GLM-5.3-Flash (max) (49.6)
- #250 Grok-2 (49.6)
- #251 ollama/qwen2.5:32b-instruct-q8_0 (49.6)
- #252 qwen3.5-4b (none) (49.6)
More from xAI
- Grok 4.6 (xhigh) (#34, 53.6)
- grok-4-0709 (#38, 53.2)
- Grok 4 (#39, 53.2)
- Grok 4.7 (xhigh) (#41, 53.1)
- Grok 4.5 (high) (#85, 52.0)
- Grok 4.20 (Reasoning) (#136, 51.2)
Compare it on the live leaderboard → · How the aggregate score works