VerdictPal
Join

Models / xAI

Grok 4.20 Reasoning

Grok 4 · Released 2026-03-05

xAI's March 2026 reasoning Grok. Vals Index 39.11 (#15). Reasoning-first variant of the Grok 4 family.

xAI reasoning route when Grok 4.3 is too expensive or too general.

#20 of 21 on Vals Index · current snapshot
39 Vals Index · 13 benchmark rows · 4 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite57
Context256K
Input / 1M—
Output / 1M—
Knowledge cutoff2026-01-01
Output speed227 t/s
TTFT (default API)16.81s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
91
Coding
46
Agentic
93
Long context
58

72average across 4 tested families

Benchmark placements

Where Grok 4.20 Reasoning places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
τ³-Bench BankingAgentic#3/ 6493Artificial Analysis2026-07-21
IFBenchReasoning#4/ 5781Artificial Analysis2026-07-21
AA output speedPerformance#7/ 69Artificial Analysis2026-07-21
AA time to first tokenPerformance#15/ 69Artificial Analysis2026-07-21
GPQA DiamondReasoning#18/ 6991Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#32/ 6532Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#41/ 6737Artificial Analysis2026-07-21
SciCodeCoding#41/ 6546Artificial Analysis2026-07-21
AA-LCRLong context#49/ 6458Artificial Analysis2026-07-21
Terminal-BenchAgentic#54/ 6938Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where Grok 4.20 Reasoning shows up, it's highlighted.

ReasoningGrok 4.20 Reasoning · 91
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingGrok 4.20 Reasoning · 46
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticGrok 4.20 Reasoning · 93
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextGrok 4.20 Reasoning · 58
  1. 1
    Kimi K3
    75
  2. 2
    GPT-5.1
    75
  3. 3
    GPT-5.6 Sol
    74
Newest receipts

Benchmark rows added to the public ledger for Grok 4.20 Reasoning in the last 120 days. Older rows live in the full table below.

13 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context58%Artificial Analysis2
AA output speedPerformance227 t/sArtificial Analysis2
AA time to first tokenPerformance16.81sArtificial Analysis2
Artificial Analysis Intelligence IndexReasoning37%Artificial Analysis2
GPQA DiamondReasoning91.1%Artificial Analysis2
Humanity's Last ExamReasoning32.2%Artificial Analysis2
IFBenchReasoning81.22%Artificial Analysis2
SciCodeCoding45.6%Artificial Analysis2
τ³-Bench BankingAgentic92.98%Artificial Analysis2
Terminal-BenchAgentic37.88%Artificial Analysis2
Vals IndexAgentic39.11%Vals AI — Vals Index49
ARC-AGIReasoning53.3%ARC Prize leaderboard69
LiveBenchReasoning56.4%LiveBench leaderboard99
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning91
Coding46
Agentic93
Long context58
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMiniMaxxAIMetaZhipuXiaomiMeituanCohere

Full benchmark ledger

Every catalog benchmark for Grok 4.20 Reasoning. Scores link to the original source; gaps mean no public row exists yet.

14 sourced rows

13 of 32 catalog benchmarks have a sourced row for Grok 4.20 Reasoning.

Reasoning

Coding

Agentic

Long context

Performance

Sources