VerdictPal
Join

Models / xAI

Grok 4.5

Grok · Released 2026-07-08

xAI's July 2026 frontier release. Artificial Analysis Intelligence Index 53.8 (high-effort setting) — above Grok 4.3 and in the current top tier with Claude Opus 4.8 and GPT-5.5. API list price $2 / $6 per 1M tokens.

Frontier xAI model for long-context and freshness-sensitive work. Prefer sourced ledger rows over marketing claims; AA scores sync from the public snapshot.

#2 of 51 on AA Index · current snapshot
54 AA Index · 9 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite60
Context2M
Input / 1M$2.0
Output / 1M$6.0
Knowledge cutoffopen
Output speed77 t/s
TTFT (default API)9.69s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
93
Coding
54
Agentic
82
Long context
68

74average across 4 tested families

Editor's note
Card scaffolded from the 2026-07-09 AA snapshot (slug grok-4-5, high effort). Editorial depth and failure modes still need a desk pass — scores below are ledger-backed, not hands-on claims.
Benchmark placements

Where Grok 4.5 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
GPQA DiamondReasoning#6/ 6993Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#7/ 6754Artificial Analysis2026-07-21
Terminal-BenchAgentic#8/ 6982Artificial Analysis2026-07-21
SciCodeCoding#11/ 6554Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#13/ 6540Artificial Analysis2026-07-21
AA time to first tokenPerformance#20/ 69Artificial Analysis2026-07-21
τ³-Bench BankingAgentic#22/ 6433Artificial Analysis2026-07-21
AA-LCRLong context#30/ 6468Artificial Analysis2026-07-21
AA output speedPerformance#33/ 69Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where Grok 4.5 shows up, it's highlighted.

ReasoningGrok 4.5 · 93
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingGrok 4.5 · 54
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticGrok 4.5 · 82
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextGrok 4.5 · 68
  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 4.5 in the last 120 days. Older rows live in the full table below.

9 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context67.67%Artificial Analysis2
AA output speedPerformance77 t/sArtificial Analysis2
AA time to first tokenPerformance9.69sArtificial Analysis2
Artificial Analysis Intelligence IndexReasoning53.8%Artificial Analysis2
GPQA DiamondReasoning93.1%Artificial Analysis2
Humanity's Last ExamReasoning40.3%Artificial Analysis2
SciCodeCoding54.1%Artificial Analysis2
τ³-Bench BankingAgentic32.58%Artificial Analysis2
Terminal-BenchAgentic81.65%Artificial Analysis2
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.

Reasoning93
Coding54
Agentic82
Long context68
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 4.5. Scores link to the original source; gaps mean no public row exists yet.

9 sourced rows

9 of 32 catalog benchmarks have a sourced row for Grok 4.5.

Reasoning

Coding

Agentic

Long context

Performance

Changelog
  1. Grok 4.5 appears on Artificial Analysis (high-effort index 53.8).
Sources