DeepSeek V4.1-Flash and GPT-6 Luna are two of the cheapest 1M-context models on the market. Luna has the lower base rates; DeepSeek has a much cheaper cache hit but doubles its price during weekday peak hours.
GPT-6 Luna is cheaper on fresh input and output ($0.10/$0.50 vs $0.15/$0.60 off-peak), and the gap widens sharply at DeepSeek's peak rate of $0.30/$1.20. DeepSeek wins only on near-total cache reuse ($0.003 vs $0.01) or on off-peak prompts above 272K tokens.
| DeepSeek V4.1-Flash | GPT-6 Luna | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Tier | Budget | Budget |
| Input / 1M tokens | $0.15 | $0.10 |
| Output / 1M tokens | $0.60 | $0.50 |
| Cache hit / 1M | $0.003 | $0.01 |
| Cache write / 1M | — | $0.125 |
| Context window | 1M | 1M |
| Notes | Released Sep 10, 2026, replacing V4-Flash & V4-Flash-Vision-Exp at a lower price; off-peak rate shown. Peak (01:00-04:00 & 06:00-10:00 UTC weekdays) is 2x: $0.30/$1.20 | New (Sep 29, 2026 DevDay), fastest & cheapest GPT-6-family tier; >272K tokens: $0.20/$0.75, cache $0.02 |
GPT-6 Luna is 33% cheaper on input tokens ($0.10 vs $0.15 per 1M). GPT-6 Luna is 17% cheaper on output tokens ($0.50 vs $0.60 per 1M).
| Workload | DeepSeek V4.1-Flash | GPT-6 Luna | Difference |
|---|---|---|---|
| Support chatbot 10,000 conversations × (2,000 input + 400 output tokens) | $5.40 | $4.00 | GPT-6 Luna saves $1.40 |
| Document Q&A 1,000 questions × (30,000 input + 800 output tokens) | $4.98 | $3.40 | GPT-6 Luna saves $1.58 |
| Coding agent session 4M input tokens (90% cache hits) + 150,000 output tokens | $0.16 | $0.15 | GPT-6 Luna saves $0.01 |
Standard rates per 1M tokens from data on this site. The coding-agent row assumes 90% of input is read from the prompt cache (full input price where a model has no cache rate). Cache-write fees, batch discounts and long-context surcharges listed in the notes are not included. Updated November 2, 2026.
your requests reuse a long cached prefix almost every time, you send prompts larger than 272K tokens, or you can schedule batch work outside DeepSeek's peak windows.
you want the lowest base rate with no time-of-day pricing, your traffic is concentrated in European mornings, or you need the OpenAI SDK and Batch API.
DeepSeek's cache hit costs $0.003 per 1M off-peak — about a third of Luna's $0.01 — and DeepSeek's pricing page lists no separate charge for writing the cache, while Luna bills $0.125 per 1M for writes. At 90% cache hits the two land close together, as the coding-agent row shows; DeepSeek only pulls ahead once hit rates pass roughly 95% at that output volume, typical of agents re-sending a long, unchanged context.
On weekdays from 01:00–04:00 and 06:00–10:00 UTC, V4.1-Flash costs $0.30/$1.20 per 1M; weekends and Chinese public holidays are off-peak. Luna has no peak pricing but steps up to $0.20/$0.75 above 272K input tokens. DeepSeek lists one rate for its full 1M window, so on very long off-peak prompts V4.1-Flash is the cheaper choice on input.
OpenAI's Batch API halves Luna's rates, which keeps Luna ahead for offline jobs. V4.1-Flash was released Sep 10, 2026, replacing V4-Flash and V4-Flash-Vision-Exp at a lower price, so older DeepSeek integrations should move to it.
Not on base rates: Luna is $0.10/$0.50 per 1M versus $0.15/$0.60 off-peak for DeepSeek. DeepSeek is cheaper on cache hits and, off-peak, on prompts over 272K tokens.
01:00–04:00 and 06:00–10:00 UTC on weekdays, excluding Chinese public holidays, when prices double.
Yes, above 272K input tokens it bills $0.20 input, $0.75 output and $0.02 cache hits per 1M.
How we choose: prices come from each vendor's official pricing page and are re-checked every week. We take no affiliate fees, so rankings aren't for sale.