OpenRouter launches the Fusion API, matching the performance of Fable 5 at half the cost, while Anthropic pauses models based on OpenRouter. On June 12, the company introduced Fusion—a server-side API that can send prompts in parallel to multiple AI models, then use review models and a synthesizer to merge the results into a single answer. In Perplexity’s DRACO benchmark, a budget panel combined Gemini 3 Flash with the open-source models Kimi K2.6 and DeepSeek V4 Pro, and synthesized it using Claude Opus 4.8 to achieve 64.7% performance—only about 1% lower than Fable 5, with costs cut by roughly half, and outperforming using GPT-5.5 (60%) and Opus 4.8 (58.8%) alone. This timing coincides with a U.S. export control order that forces Anthropic to suspend Fable 5 and Mythos 5 for foreign nationals worldwide. OpenRouter positions Fusion as “Fable-level intelligence, half the price” to make up for the gap caused by the suspension.

View Original
GateNews
OpenRouter Launches Fusion API Matching Fable 5 Performance at Half Cost as Anthropic Suspends Model
According to OpenRouter, the company launched Fusion on June 12, a server-side API that sends prompts to multiple AI models in parallel, then uses a judge model and synthesizer to merge results into a single answer. On Perplexity's DRACO benchmark, a budget panel combining Gemini 3 Flash with open-s
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned