Claude Opus 4.8 Fast

• 🆕 Speed-optimized variant of Anthropic's flagship Claude Opus 4.8 model
• ⚡ Fast path prioritizes higher output throughput and lower latency
• 📏 Same 1M-token context window and 128k max output tokens
• 🧠 Hybrid reasoning with adaptive thinking, defaulting to high effort
• 🔧 Tuned for long-horizon agentic coding and tool use
• 👁️ Accepts text and image inputs, plus function calling and web search
• 🔒 Anthropic reports improved honesty and fewer unsupported claims
• 🏢 Premium-priced fast tier aimed at latency-sensitive agentic pipelines

Anthropic

Anthropic is an American artificial intelligence company headquartered in San Francisco, focused on developing large language models under the Claude family. Founded with a strong emphasis on AI safety research, the company has become one of the most…

Claude Opus 4.8 Fast, released on May 28, 2026, is a speed-optimized configuration of Claude Opus 4.8, which Anthropic describes as its most capable generally available Opus model for coding and agents. It runs the identical underlying model — the same 1M-token context window, 128k maximum output tokens, and adaptive thinking — but routes inference through a higher-throughput "fast" path.

The fast mode's benefit centers on output token generation rather than time to first token, making it suited to live coding sessions and real-time agentic loops, where it trades premium pricing for reduced latency. Standard Opus 4.8 offers the same intelligence at lower cost for batch and less time-sensitive workloads.

Relative to its predecessor, Anthropic says Opus 4.8 targets behavioral improvements over Claude Opus 4.7 in long-horizon agentic coding — including better long-context handling, fewer compactions, improved compaction recovery, and more reliable tool triggering, addressing a reported tendency in Opus 4.7 to skip required tool calls. Anthropic also highlights honesty gains, noting early testers found the model more likely to flag uncertainty and less likely to make unsupported claims.