GitHub Copilot Pro

LLM 3.7/10

Unlimited code completions + premium requests (legacy 300/mo, now usage-based AI Credits, 1 credit = $0.01) across GPT-5.5/5.4/Codex, Claude Opus 4.8/Sonnet 5, Gemini 3.1/3.5, and Kimi K2.7 Code (GA); agent mode in VS Code/JetBrains/CLI

Requests / month: 300

GitHub Copilot Pro scores 3.7/10 for free value in LLM. The main constraint: Paid — $10/mo (Pro+ $39/mo with 1,500 premium requests). Unlimited code completions + premium requests (legacy 300/mo, now usage-based AI Credits, 1 credit = $0.01) across GPT-5.5/5.4/Codex, Claude Opus 4.8/Sonnet 5, Gemini 3.1/3.5, and Kimi K2.7 Code (GA); agent mode in VS Code/JetBrains/CLI.

Category LLM Kind Paid plan Caveats No general OpenAI/Anthropic-compatible endpoint for external agents (unofficial copilot-api proxies violate ToS); agent mode burns several premium requests per session — heavy use exhausts Pro in ~a week

Free value score

3.7 / 10

Show the math

Model Quality 7.5/10

value = 42 artificial_analysis_score

weighted score = 0.2 * 7.5/10

Requests 1/10

value = 300 req/mo

weighted score = 0.1 * 1/10

Access 10/10

value = 10/10

score = 10/10

Price Efficiency 3/10

value = 0.1 value/$

weighted score = 0.7 * 3/10

Key metrics

Kimi K2.7 Code 42 Best model

300 R/mo Requests/month

0.1x Value ratio

What you get

Paid — $10/mo (Pro+ $39/mo with 1,500 premium requests). Unlimited code completions + premium requests (legacy 300/mo, now usage-based AI Credits, 1 credit = $0.01) across GPT-5.5/5.4/Codex, Claude Opus 4.8/Sonnet 5, Gemini 3.1/3.5, and Kimi K2.7 Code (GA); agent mode in VS Code/JetBrains/CLI

Cheapest access to frontier closed models + Kimi K2.7 Code inside an IDE/CLI; poor fit for external custom agents (no raw API)

Gotchas

No gotchas documented yet. Contribute one if you know a constraint we missed.

Sources

"Copilot moved to usage-based AI Credits June 1 2026 (1 credit = $0.01); Pro legacy allowance 300 premium requests/mo, overage $0.04/request"
https://docs.github.com/en/copilot/reference/copilot-billing/models-and-pricing
"Kimi-K2.7-Code (Moonshot AI) generally available in Copilot; Grok Code Fast 1 retired"
https://docs.github.com/copilot/reference/ai-models/supported-models

paid value llm-providers

Open service ← All LLM

Alternatives

Same category · similar quota

Kimi Code ModeratoKimi Code Moderato · paid · Paid plan

3.7

Kimi K2.7 Code + K2.7 HighSpeed (~180 tok/s) + K2.6; ~300-1,200 requests/5h with weekly quota, up to 30 concurrent; OpenAI- AND Anthropic-compatible endpoints (Claude Code, OpenCode, OpenClaw, custom agents)

0.05×42 AA0.1×

May be outdatedChecked 23d ago

Paid only

GitHub ModelsGitHub Models · freemium · Hosted multi-model LLM API

3.8

Every GitHub account gets free, rate-limited API access to a large catalog of frontier and open models (GPT-5/4.1, Llama, DeepSeek, Grok, Mistral, Cohere, Phi) for prototyping.

20.1 AA1/min

May be outdatedChecked 28d ago

NoncomEvalEmbeddings

Cohere Trial API KeyCohere Trial API Key · freemium · First-party hosted LLM API

4.0

Cohere auto-issues a free, rate-limited Trial API key on signup giving access to Command A/R/R+, Aya, plus Embed and Rerank models for noncom use.

29.3 AA20/min

May be outdatedChecked 28d ago

EmbeddingsNoncom