V4 Flash Scores 82.7 on Terminal Bench at Just $0.28/M Tokens
jiayuan_jy · x · 2026-07-31
A user highlighted the impressive performance of the newly updated V4 Flash model on the Terminal Bench. It achieved a high score of 82.7 at a remarkably low cost of $0.28 per million tokens. For context, GLM 5.2 (which is 4 times larger) scored 81%, and Kimi K3 (12-13 times larger) scored 88.3%. The results showcase the model's extraordinary price-to-performance ratio.
More from Models
- DeepSeek-V4 Flash Scores 50 on AA Index, Beating Price-Cut GPT 5.6 in Cost-Efficiency — zainhas · 2026-07-31
- Specific Prompt Triggers 'Existential Crisis' and Role Confusion in Claude Opus 5 — Any-Reputation8118 · 2026-07-31
- Gemini V4 Rumored to Train on Tens of Trillions of Tokens for Multimodal — teortaxesTex · 2026-07-31
- DeepSeek V4-Flash Cracks Complex Russian Joke That Trips Up Other LLMs — teortaxesTex · 2026-07-31
- Model Selection is Becoming Org Design: Structuring AI Workflows — every · 2026-07-31
- DeepSeek-V4-Flash Benchmarks Leak: Higher IQ Density but Questionable Pricing — teortaxesTex · 2026-07-31