← All models

Best Model for Aider

Aider works with virtually any LLM and is famous for its code-editing benchmarks, which makes model choice the single biggest lever on its quality. Claude Opus 4.8 and GPT-5.5 sit at the top; Claude Sonnet 4.6 is the value-quality sweet spot; and DeepSeek V4 Pro, GLM 5.2, and Kimi K2.7 Code give Aider excellent results at open-weight prices.

Read the full Aider review → · updated 2026-06-22
#1Claude Opus 4.8Anthropic · $11/1M · 1M ctx

Top of the code-editing benchmarks — the highest-quality option for Aider's diff-based edits.

Overall 94/100 · Value 41 · 79% community winFull breakdown →
#2Claude Sonnet 4.6Anthropic · $6.60/1M · 1M ctx

The value-quality sweet spot — most of Opus's editing ability for far less.

Overall 91/100 · Value 46 · 78% community winFull breakdown →
#3GPT-5.5OpenAI · $13/1M · 1.1M ctx

OpenAI's flagship — neck-and-neck with Opus on Aider's hardest edits.

Overall 93/100 · Value 40 · 79% community winFull breakdown →
#4DeepSeek V4 ProDeepSeek · $0.57/1M · 1.0M ctx

The value champion — top open-weight benchmark scores at a fraction of the cost.

Overall 88/100 · Value 83 · 77% community winFull breakdown →
#5Gemini 3.1 ProGoogle · $5.00/1M · 1.0M ctx

The pick for very large repos thanks to its 1M-token context.

Overall 89/100 · Value 49 · 77% community winFull breakdown →
#6GLM 5.2Z.AI · $1.61/1M · 1.0M ctx

A strong open-weight coder with a 1M context — great value for Aider.

Overall 89/100 · Value 65 · 77% community winFull breakdown →

FAQ

What is the best model for Aider?
Claude Opus 4.8 and GPT-5.5 top Aider's code-editing benchmarks; Claude Sonnet 4.6 is the best value-quality balance. For open weights, DeepSeek V4 Pro and GLM 5.2 are excellent. Aider supports virtually any model, so you can pick freely.
Does Aider's edit quality depend on the model?
Almost entirely — Aider is a thin, disciplined layer over the model, so a stronger model directly means better edits. Aider's own leaderboard tracks this, and it tracks closely with the picks here.

The model is half the story — the agent is the other half

The model picks the moves; the agent runs the loop, the tools, and the guardrails. Once you've chosen a model, see which agent gets the most out of it.

Compare AI agents →