LiteLLM 1.108.2 released
An LTS patch release that fixes request routing, API compatibility, and error handling across multiple providers.
What changed in 1.108.2
- OpenAI deployments with empty API keys no longer send admin credentials as bearer token.
- AWS Bedrock now honors service tier settings and preserves Claude 4.5 beta features in invoke calls.
- Fixed JSON response format errors, base64 PDF handling for Claude, and thinking drops in schema calls.
- Chat completions with guardrails, prompt caching, and affinity filtering now route requests correctly.
- Audio transcription, MCP hosted tools, and cost tracking errors resolved.
Summarised by AgentGid from BerriAI's official release notes on GitHub, as reported there; read the original before relying on it. Price, status and release dates come from our daily data.
LiteLLM at a glance
| What it is | An AI gateway and Python SDK that calls 100+ model APIs through one OpenAI-style interface, with spend tracking, budgets, rate limits, fallbacks, guardrails and logging. The same gateway also serves A2A agents and MCP t… |
|---|---|
| Latest version | 1.106.10, Nov 14, 2026. This page is about an earlier release. |
| Release pace | 41 stable releases in the last 30 days; 1.108.2 came 1 day after the previous one |
| Price | Free (OSS). The SDK and proxy are open source and free. An enterprise edition with SSO, audit logs and support is sold separately. Official pricing ↗ (checked Nov 5, 2026) |
| Downloads, last 7 days | 23.8M up 5.3% |
| Gid Score | 76 #1 in Agent tooling for cost, context and memory |
