Models on CoralBricks
3 public models behind one OpenAI-compatible API at https://inference.coralbricks.ai/v1, each with the same API key. Rates are the Standard plan, in USD.
DeepSeek V4.1 Flash Value
- Model ID
deepseek-v4.1-flash-value- Context window
- 1M tokens
- Input
- $0.005 per 1M tokens
- Cached input
- $0.003 per 1M tokens
- Output
- $0.57 per 1M tokens
- Cost against DeepSeek's V4.1 Flash price
- 75% below
- Decode speed target
- >50 tok/s
- Uptime, last 30 days
- 100.00%
- Decode speed, p50 over 24 hours
- 136 tok/s
- Time to first token, p50 over 24 hours
- 710 ms
DeepSeek V4.1 Flash Standard
- Model ID
deepseek-v4.1-flash- Context window
- 1M tokens
- Input
- $0.19 per 1M tokens
- Cached input
- $0.008 per 1M tokens
- Output
- $0.60 per 1M tokens
- Cost against DeepSeek's V4.1 Flash price
- 25% below
- Decode speed target
- >100 tok/s
- Uptime, last 30 days
- 100.00%
- Decode speed, p50 over 24 hours
- 140 tok/s
- Time to first token, p50 over 24 hours
- 710 ms
DeepSeek V4.1 Flash Fast
- Model ID
deepseek-v4.1-flash-fast- Context window
- 1M tokens
- Input
- $0.45 per 1M tokens
- Cached input
- $0.009 per 1M tokens
- Output
- $1.80 per 1M tokens
- Cost against DeepSeek's V4.1 Flash price
- 1.5×
- Decode speed target
- >300 tok/s
- Uptime, last 30 days
- 99.95%
- Decode speed, p50 over 24 hours
- 441 tok/s
- Time to first token, p50 over 24 hours
- 350 ms
Uptime, decode speed and time to first token as measured at Nov 20, 2026, 12:29 PM PT, when this page was built. Current figures: the models page (live speed and uptime) and the status page (status, 30-day uptime, incident history).
Cost compares a day of our own traffic mix with DeepSeek's published peak-hour V4.1 Flash price; individual token categories vary.