DeepSeek deprecates `deepseek-chat` and `deepseek-reasoner` for V4-Flash and V4-Pro
thejasminejade · x · 2026-07-25
DeepSeek has deprecated deepseek-chat and deepseek-reasoner in favor of its V4-Flash and V4-Pro naming.
The post warns that code may auto-map for now, but teams with serious or dependent workflows should update their integrations to the new model names.
More from Models
- Anthropic launches Claude Opus 5 at the same $5/$25 token pricing as Opus 4.8 — ctjlewis · 2026-07-25
- Claude Opus 5 tops a prompt-injection robustness chart and drops attacks near zero — EricBuess · 2026-07-25
- Users joke Claude Opus 5 feels like GPT-5 as Anthropic touts SOTA scores — ChengleiSi · 2026-07-25
- Claude Opus 5 works well for coding, but it breaks Compound Engineering workflows — danshipper · 2026-07-25
- Claude Opus 5 is being pitched as cheaper per task while matching stronger models — EricBuess · 2026-07-25
- Anthropic launches Claude Opus 5 as its strongest model for biology and chemistry — gabepgomes · 2026-07-25