Best API & Subscription for Aider

To run Aider cost-efficiently, set OPENAI_API_BASE=https://api.stdcmpt.com/v1 and run aider --model openai/standardcompute: Standard Compute's router sends each request to the cheapest model that can do it (Claude Fable 5.1 or GPT-6 Astra only when it counts), on a plan from $19 a month with a $20 compute budget. Usage-based access through OpenRouter can suit occasional tasks; choose around edit quality, repository context and how often you use Aider, and verify editing behavior on your own repository.

Light or occasional use

Use only the repository context you need and compare a lower-cost model on a real edit. If your setup supports a separate weak model, reserve it for suitable auxiliary tasks. Compare measured usage with any subscription allowance.

Daily use, cost-optimized

Standard Compute routes each Aider call to the cheapest model that can do it, and for every model we vet providers for price, uptime and speed, with automatic fallback. Opt-in Europe or US Compliance rules apply per key, team or company, and we don't train on your prompts or store them. The monthly plan bounds the bill, but Aider's calls still consume its compute allowance: test a representative repository task and review usage before choosing a tier.

API and subscription options for Aider

Prices and allowances can change. OpenCode Go prices were checked against its linked public pages on October 13, 2026, DevPass prices on September 5, 2026; other entries link to current plans without quoting an unverified price. Spot an error? Tell us.

Aider questions

Is there a fixed monthly, OpenAI-compatible API for Aider with GPT-5.6 Terra?

Yes. Set OPENAI_API_BASE=https://api.stdcmpt.com/v1 and run aider --model openai/openai/gpt-5.6-terra: Standard Compute serves Terra with its 1.05M-token context on every plan from $19 a month, with no OpenAI usage tier. Terra is OpenAI's closed model, so there is no third-party quantized copy to land on. OpenAI lists Terra at $2 / $12 per million tokens and the plan's budget is drawn at Terra's rate, so a full-time Terra default wants Standard ($89 a month, $95 budget) or larger. Claude Sonnet 5.5 and Gemini 3.1 Pro Preview are long-context peers on the same key.

Where can I run Qwen 3.8 27B for Aider without DashScope's free-tier throttling?

Standard Compute serves qwen/qwen3.8-27b on every plan: set OPENAI_API_BASE=https://api.stdcmpt.com/v1 and run aider --model openai/qwen/qwen3.8-27b. There is no daily quota that resets on Beijing time, only a monthly compute budget and a per-key anti-abuse rate ceiling, so a European morning session starts with the same limits as any other hour. Plans start at $19 a month. Alibaba lists 27B at $0.50 / $3 per million tokens, and its free 1M-token quota is valid for 90 days.

I'm growing from a freelancer into a 3-person studio on Aider and Cline. What is the simplest flat multi-seat setup where I can see who spent what?

A free company workspace with one key per person. Each key gets its own monthly plan, approved and paid by you as admin in one place: three keys on Economy cost 3 x $39 = $117 a month (USD) with $41 of compute each, and anyone who needs more moves to Standard at $89 with $95. Aider takes a key through OPENAI_API_BASE=https://api.stdcmpt.com/v1 with --model openai/standardcompute, and Cline through its built-in Standard Compute provider (OpenAI Compatible on 4.1.24 and older) at the same URL. Usage and the request log are per key and show the model, tokens and latency of each request, so you see who spent what, and a key that reaches its budget stops with HTTP 402 without touching the other two.