Blog & Changelog · Aevon

Updates, guides,
and what's new.

Product updates, release notes, and developer guides from the Aevon team.

● Featured · Jun 7, 2026

Introducing Aevon — one key,
every frontier model.

We built Aevon because we were tired of managing 6 different API keys and billing dashboards. One endpoint, one key, 30+ models — that's the whole idea.

Jun 7, 2026 · 4 min read Read Post →
v1.0.3
Jun 7
2026

Qwen3 & Llama 3.3 added.

  • NEWAdded qwen3-235b (Alibaba) — 32K context
  • NEWAdded llama-3.3-70b (Meta) — free tier, 128K context
  • IMPReduced P50 latency by 8ms across all routes
  • FIXFixed streaming disconnect on long responses (>8K tokens)
v1.0.2
Jun 1
2026

Dashboard analytics & usage logs.

  • NEWUsage analytics page — spending breakdown per model
  • NEWRequest logs with status codes, tokens, cost per call
  • NEWAPI key reveal/hide toggle in dashboard
  • IMPImproved credit estimator accuracy
  • FIXRate limit headers now correctly returned on 429
v1.0.1
May 28
2026

Gemini 2.0 Pro & stability fixes.

  • NEWAdded gemini-2.0-pro (Google) — 1M context window
  • IMPDeepSeek R1 routing optimized — 340ms faster avg
  • FIXAuth token expiry edge case fixed
  • FIXCredit deduction precision corrected to 6 decimal places
v1.0.0
May 20
2026

Initial launch. Aevon is live.

  • NEWOpenAI-compatible gateway — /v1/chat/completions
  • NEW10 models at launch: Claude, GPT-4o, DeepSeek, Qwen, GLM
  • NEW$100 free credits on GitHub signup
  • NEWDashboard with API key management
  • NEWReal-time usage metering