One $2,618 prompt: what Claude's newest Opus model built overnight
altryne · x · 2026-11-11
Peter Gostev ran an extreme test: a single $2,618 prompt that let Anthropic's newest Opus-tier model (dubbed Fable 5.2 Max) autonomously build a project overnight. In his video he walks through dozens of prompts.
- Overall: the model improves on many fronts, but not across the board
- Verdict: in some cases it still makes sense to stick with Opus rather than switching unconditionally
Related event: Tester Spends $2,618 on a Single Prompt with Fable 5.2(2 posts)→
More from coding & agent
- Dev demos agent with 'durable object brain' plus on-demand computer, local or remote — samgoodwin89 · 2026-11-11
- Right-sizing intelligence per role: frontier reasoning at the control plane, adversarial verifiers at the gate — clmt · 2026-11-11
- inkbox.ai Raises $5.3M Led by Peak XV to Give AI Agents Phone Numbers and Identities — cneuralnetwork · 2026-11-11
- Claude can now host your full site and act on what's on screen via MCP Apps — ClaudeDevs · 2026-11-11
- MCP Apps support lands in public preview as spec is standardized with community — ClaudeDevs · 2026-11-11
- Claude expands MCP Apps support: full apps and embedded sites now run inside Claude — ClaudeDevs · 2026-11-11