LLM ReferenceLLM Reference

Grok-3

grok-3

DeprecatedResearched 2d ago

Last refreshed 2026-05-16. Next refresh: weekly.

ProprietaryMultimodalCodingRAGAgentsLong contextVisionClassificationJSON / Tool use

Grok-3 is a legacy integration reference; evaluate Grok 4.3 before starting new work.

Decision context: Coding task fit, 4 tracked provider routes, and research from 2026-05-16.

Use it for

  • Teams maintaining an existing integration
  • Workloads that can use a 1M context window
  • Buyers comparing 3 tracked provider routes

Do not use it for

  • New production launches

Cheapest output

$2.40

Chutes AI per 1M tokens

Provider routes

4

Tracked API hosts

Quality / dollar

Unknown

No output-token price in the ladder

Freshness

2026-05-16

Researched 2d ago

fresh

This API model is marked deprecated. Use Grok 4.3 as the replacement candidate before sending new production traffic.

Top use-case fit

Coding

1 relevant benchmark in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Agents

Included by capability and metadata signals in the decision map.

Provider price ladder

ProviderInput / 1MOutput / 1MRoute
Chutes AI$0.800$2.40
Serverless
Microsoft Foundry$3.00$15.00
Serverless
OpenRouter$3.00$15.00
Serverless

Benchmark peer barsfor Coding

Aider PolyglotRank 18 of 29

Migration checks

No linked migration route is available for this model yet.

About

Grok-3 is a 2025 xAI model confirmed for May 15, 2026 retirement; use Grok 4.3 for current API work.

Grok-3 has a 1M-token context window.

Grok-3 input tokens at $0.8/1M, output at $2.4/1M.

Capabilities

VisionMultimodalReasoningFunction CallingTool UseStructured Outputs

Benchmark Scores(4)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
Aider Polyglot53.32026-04https://aider.chat/docs/leaderboards
Chatbot Arena1405.0https://lmarena.ai
Massive Multi-discipline Multimodal Understanding78.0https://mmmu-benchmark.github.io/
MMLU PRO79.9https://huggingface.co/spaces/TIGER-Lab/MMLU-Pro

Rankings

Specifications

FamilyGrok 3
Released2025-04-09
Parameters1B
Context1M
Knowledge cutoff2025-04

Created by

Ethical AI for universal truth-seeking

San Francisco, California, United States
Founded 2023
Website