feat: add Avian as a new API provider

[avianion]

Summary

Adds Avian as a new OpenAI-compatible API provider with four models:

• DeepSeek V3.2 — 164K context, 65K output
• Kimi K2.5 — 131K context, 8K output
• GLM-5 — 131K context, 16K output
• MiniMax M2.5 — 1M context, 1M output

Changes

• packages/opencode/src/provider/models.ts — Add Avian provider definition with model metadata (context limits, capabilities, pricing)
• packages/opencode/src/provider/provider.ts — Add Avian provider loader (API key detection via env var, auth store, or config)

Follows the same pattern as other OpenAI-compatible providers (uses @ai-sdk/openai-compatible). Only 2 files changed, 87 lines added.

Test plan

• Set AVIAN_API_KEY env var and verify Avian appears in provider list
• Verify model selection shows all 4 models with correct names
• Test chat completion with DeepSeek V3.2 model

🤖 Generated with Claude Code

[avianion]

Hi @chrarnoldus @kevinvandijk — tagging you as you've been actively reviewing and merging provider-related PRs. Would appreciate a review when you have a chance. Thanks!

[kilo-code-bot]

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Files Reviewed (4 files)

• packages/kilo-docs/lychee.toml
• packages/kilo-docs/source-links.md
• packages/opencode/src/provider/models.ts
• packages/opencode/src/provider/provider.ts

---

_{Reviewed by gpt-5.5-20260423 · 808,519 tokens}

[avianion]

Thanks for the review! The 1,000,000 output token limit for MiniMax M2.5 is correct — MiniMax M2.5 supports up to 1M tokens for both input context and output. This is a distinguishing feature of the model (one of the largest output windows available). No change needed here.

[avianion]

Friendly follow-up — this PR is still active and ready for review. Would appreciate a look when you get a chance! cc @marius-kilocode @iscekic

[avianion]

Friendly follow-up — this PR is still active and ready for review. All feedback has been addressed. Would appreciate a look when you get a chance! cc @marius-kilocode @iscekic

[avianion]

Hey @thdxr @adamdotdevin — friendly follow-up on this PR. Avian is an OpenAI-compatible inference provider that's already live and powering apps like ISEKAI ZERO. This is a lightweight integration (standard OpenAI-compatible endpoint) and we're happy to address any feedback or make adjustments. Would love to get this merged if you have a moment to review. Thanks!

[chrarnoldus]

@avianion could you resolve the merge conflicts?

[avianion]

Resolved merge conflicts — merged latest upstream main. The conflict was in packages/opencode/src/provider/models.ts where the Apertis provider was added on main while this PR adds the Avian provider. Both are now included. Ready for review.

[marius-kilocode]

@chrarnoldus could you take another look?

[avianion]

Resolved merge conflicts with latest upstream.