Skip to content

xAI · Chat

Grok 4.5

xAI 🇺🇸ChatGrok familyOne of its sources says the weights aren't publicly available.The maker lists this model as deprecated (19 Oct 2026).
From Neosantara's listing:
Grok 4.5 is a frontier reasoning model from xAI built for coding, agentic tasks, and knowledge work. It accepts text, image, and PDF input with selectable low, medium, and high reasoning depth, a 500K token context window, and tool use spanning function calling, structured outputs, and search.
llminfo IQ
136.6
Maker's price
$3.00 blended / 1M
Max context
2M
Listings
54

Evidence and sources

llminfo IQ 136.6, ranked among chat models. From 312 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 117.4, 95% interval 109–125. From 29 benchmark families across 15 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (56.9). Cheaper per task than its benchmark results predict. From 50 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsTerse (37 ±9). Generates fewer tokens than models with similar coding results. Measured at high effort. A fair amount of data behind it. By effort: default 46 · low 41 · medium 42 · high 37.Median 25,420 output tokens per task at high effort.How the Yap score worksReleased 8 Jul 2026Maker HQ: US (according to xAI's own website)
All rankings: #114 on OpenRouter this week #116 on OpenRouter today #117 on OpenRouter this month #350 on NanoGPT recentlySources: LMArena data, LiteLLM data, Models.dev data, AI/ML API data, Neosantara Inference data, llminfo researchPrimary data from LMArena from LiteLLM: deprecation date from Models.dev: weights availability from AI/ML API: primary mode from Neosantara Inference: description from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Trade-offs

Deprecated by its maker

The maker has already deprecated this model (19 Oct 2026).

No open weights

The weights aren't public, so you can't run it yourself.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 10 maker's prices.

Input & output

as stated by the model's providers
Text
Images
Files
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt cachingWeb search

Listings

 xAIxai/grok-4.5-latest · latestSource: LiteLLM data500K$2.00>200k $4.00cache read $0.30$6.00>200k $12.00$2.00
 Tokenmixgrok-4.5Sources: TokenMix data, TokenMix PricingNot published by the provider$1.90$5.70—
 302aigrok-4.5Source: Models.dev data500K$2.00cache read $0.50$6.00—
 Atlascloudxai/grok-4.5Source: AtlasCloud data500K$2.00cache read $0.30$6.00—
 Cometapigrok-4.5Sources: CometAPI data, CometAPI Pricing dataNot published by the provider$2.00$6.00—
 Eden AIInference by xAIxai/grok-4.5Source: Models.dev data500K$2.00cache read $0.30$6.00—
 Evolinkgrok-4.5Sources: EvoLink data, EvoLink PricingNot published by the provider$2.00$6.00—
 Fastrouterx-ai/grok-4.5Source: FastRouter data500K$2.00>200k $4.00cache read $0.30batch $1.60$6.00>200k $12.00batch $4.80—
 Kilox-ai/grok-4.5Sources: Kilo AI Gateway data, Models.dev data500K$2.00cache read $0.30$6.00—
 Llmgatewaygrok-4-5Source: LLMGateway data500K$2.00cache read $0.30$6.00—

8 listings without a price are hidden.

Included in plans

Copilot EnterprisegithubSubscription$39/moextra use billed past the allowancePlan page
Copilot ProgithubSubscription$10/moextra use billed past the allowancePlan page
Copilot Pro+githubSubscription$39/moextra use billed past the allowancePlan page
Copilot BusinessgithubSubscription$19/moextra use billed past the allowancePlan page
Copilot MaxgithubSubscription$100/moextra use billed past the allowancePlan page

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO63 related benchmarks · 63 results
Source: LMArena data · Nov 2026
BenchmarkScore
Arena ELO
Chinese
Creative Writing
Document
Show 59 more benchmarksHide 59 benchmarks
BenchmarkScore
Document Style-Controlled
English
Exclude Ties
Expert
French
German
Hard Prompts
Hard Prompts English
Industry Business and Management and Financial Operations
Industry Entertainment and Sports and Media
Industry Life and Physical and Social Science
Industry Mathematical
Industry Medicine and Healthcare
Industry Software and It Services
Industry Writing and Literature and Language
Japanese
Korean
Longer Query
Multi Turn
Non English
Polish
Russian
Search Factuality
Search Style-Controlled
Spanish
Style-Controlled
Text Factuality
Text Factuality - Chinese
Text Factuality - Creative Writing
Text Factuality - English
Text Factuality - Exclude Ties
Text Factuality - Expert
Text Factuality - French
Text Factuality - Hard Prompts
Text Factuality - Hard Prompts English
Text Factuality - Industry Business and Management and Financial Operations
Text Factuality - Industry Entertainment and Sports and Media
Text Factuality - Industry Life and Physical and Social Science
Text Factuality - Industry Mathematical
Text Factuality - Industry Medicine and Healthcare
Text Factuality - Industry Software and It Services
Text Factuality - Industry Writing and Literature and Language
Text Factuality - Longer Query
Text Factuality - Multi Turn
Text Factuality - Non English
Text Factuality - Polish
Text Factuality - Russian
Text Factuality - Spanish
Vision
Vision - Chinese
Vision - Creative Writing Vision
Vision - Diagram
Vision - English
Vision - Homework
Vision - Humor
Vision Style-Controlled

General

Coding

Arena ELOSource: LMArena data · Nov 2026
SWE-Marathon2 results
Reported by SpaceXAI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
resolution rate · pass@1
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
ValsSource: Vals AI

Agentic

GDPval-AA v22 results
Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Gemini 3.6 Flash Model Card · July 2026
Reported by Meta · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
MLE-BenchGemini 3.6 Flash Model Card · July 2026 · Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
SWE-Bench Pro3 results
Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Gemini 3.6 Flash Model Card · July 2026 · Public
Reported by SpaceXAI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Tau3 Banking5 related benchmarks · 5 results
tau3-bench v1 banking-knowledge pass^1
Source: Sierra tau-bench Leaderboard data docs · Jul 2026
submission-grok-4-5_sierra_2026-08-04, usersim-gpt-5.2, effort-high, retrieval-alltools
tau3-bench v1 banking-knowledge pass^2
Source: Sierra tau-bench Leaderboard data docs · Jul 2026
submission-grok-4-5_sierra_2026-08-04, usersim-gpt-5.2, effort-high, retrieval-alltools
tau3-bench v1 banking-knowledge pass^3
Source: Sierra tau-bench Leaderboard data docs · Jul 2026
submission-grok-4-5_sierra_2026-08-04, usersim-gpt-5.2, effort-high, retrieval-alltools
tau3-bench v1 banking-knowledge pass^4
Source: Sierra tau-bench Leaderboard data docs · Jul 2026
submission-grok-4-5_sierra_2026-08-04, usersim-gpt-5.2, effort-high, retrieval-alltools
Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Gemini 3.6 Flash Model Card · July 2026 · Terminus-2
Reported by Meta · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Reported by SpaceXAI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
xAI launch comparison table
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Reasoning

Math

Multilingual

EKI EstonianSource: Keelemudelite mõõdupuu data docs · Aug 2026

Vision

CharXiv ReasoningGemini 3.6 Flash Model Card · July 2026 · reasoning · no tools · Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.

Science

Factuality

Instruction following

OCR

Knowledge

Long context

Speech

Other

DeepSWE v1.0Datacurve eval · each provider's harness · pass@1 · Reported by SpaceXAI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
DeepSWE v1.13 results
Reported by SpaceXAI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Datacurve mini-swe-agent harness
Reported by Google DeepMind · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Gemini 3.6 Flash Model Card · July 2026
Reported by Meta · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Horangi36 related benchmarks · 36 results
Nejumi 428 related benchmarks · 28 results
PerceptionBench9 related benchmarks · 9 results
Overall accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Attribute accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Comparison accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Contextual integration accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Counting accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Depth and 3D perception accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Show 3 more resultsHide 3 results
Fine-grained recognition accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Localization accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
Visual relation accuracy
Reported by Moonshot AI / PerceptionBench authors · via BenchBoard docs
benchmark-owner evaluation · 3,000 verified questions total · open-ended short answers · unified prompts · highest available reasoning budget; level not specified · main-table result; run count not specified · GPT-oss-120B judge
SkillsBench · Vals6 related benchmarks · 6 results
With Skills · Accuracy
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · With Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
With Skills · Mean cost / test
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · With Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
With Skills · Mean duration / test
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · With Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
Without Skills · Accuracy
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · Without Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
Without Skills · Mean cost / test
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · Without Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
Without Skills · Mean duration / test
Reported by Vals AI · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Vals AI · updated 2026-09-27 · metadata version 1 · 87 tasks in trace charts · 3 trials/task/condition · per-task mean then task average · OpenHands-CLI@ · Without Skills · system=grok/grok-4.5 · provider=SpaceXAI · reasoning_effort=high · temperature=0.7 · top_p=0.95
Vals13 related benchmarks · 15 results
EMB
Source: Vals AI
CorpFin v2
Source: Vals AI
MedCode
Source: Clinical Benchmarks data docs · Sep 2026
Setting not stated
Source: Vals AI
Setting not stated
Show 8 more resultsHide 8 results
MedScribe
Source: Clinical Benchmarks data docs · Sep 2026
Setting not stated
Source: Vals AI
Setting not stated
MortgageTax
Source: Vals AI

Data credits: Data by LLM Stats · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · LMArena, CC BY 4.0 (licence) · BullshitBench by Peter Gostev, MIT (licence) · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · EQ-Bench (Samuel J. Paech) — cite paech2023eqbench · CyberGym leaderboard (Wang, Shi, He, Cai, Zhang, Song; UC Berkeley), ICLR 2026. Results evaluated and submitted by the listed teams. No licence is stated for the result data; the repository's Apache-2.0 licence covers its code only. · Keelemudelite mõõdupuu: Independent LLM leaderboard for Estonian. Eesti Keele Instituut. https://moodupuu.eki.ee, CC BY 4.0. Latest per-benchmark results selected with partial-run flags retained. (licence) · Horangi 4, Weights & Biases Korea; public publisher-selected numeric evaluation results · MineBench, MIT (licence) · Nejumi Leaderboard 4, Weights & Biases Japan; public publisher-selected numeric evaluation results · Sierra Research tau2-bench leaderboard submissions (github.com/sierra-research/tau2-bench), MIT (licence) · Clinical Benchmarks (clinicalbenchmarks.ai), CC BY 4.0; fractions converted to percentages; source documents retain their own terms (licence)

Around the web

Also known as: github_copilot/grok-4.5, gmi/x-ai/grok-4.5, grok-4-5, grok-4.5-latest, grok-4.5:web, grok-build-latest, grok/grok-4.5, perplexity/xai/grok-4.5, spacexai/grok-4.5, x-ai/grok-4-5, x-ai/grok-4.5, xai/grok-4-5, xai/grok-4.5, xai/grok-4.5-latest

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works