DeepSeek V4.1 Flash (1017) API Pricing / Price per 1M tokens
DeepSeek • API Price Comparison
Also known as: deepseek-v4.1-flash · deepseek v4.1 flash · deepseekv4.1flash · deepseekv4.1-flash · deepseekv4.1 flash · deepseek-v4.1-flash-(1017) · deepseekv4.1flash(1017) · deepseekv4.1-flash-(1017) · deepseekv4.1 flash (1017)
Updated on · Published currencies / per 1M tokens; cheapest ranked by input
| Channel | Input / 1M | Output / 1M | Cache read / 1M | Updated / verified on |
|---|---|---|---|---|
| DeepSeekOfficial | USD 0.3 | USD 1.2 | USD 0.006 | 2026-10-25 |
| OpenCode ZenLowest input | USD 0 | USD 0 | USD 0 | 2026-10-15 |
| OpenCode Go | USD 0.15 | USD 0.6 | USD 0.003 | 2026-10-15 |
| Together AI | USD 0.3 | USD 1.2 | — | 2026-10-25 |
— means the price or date is not recorded. Cross-currency ranking uses exchange rates.
DeepSeek V4.1 Flash (1017) is a large language model by DeepSeek with a 1.0M-token context window. Across the 6 API channels tracked here, input pricing ranges from $0.15–$0.30 per 1M input tokens (USD-normalized). The official channel (DeepSeek) charges $0.30 per 1M input tokens and $1.20 per 1M output tokens. Channels reachable from mainland China without a VPN include DeepSeek. It is also available through subscription plans such as Go, Go Plus. Its Agent Arena net-improvement score is approximately 3.8%. Prices are collected daily from official provider pages and every tracked channel, then cross-checked by a data-accuracy audit.
Model performance benchmarks
Scores retain their original benchmark, version and metric. We do not combine them into a site-wide score; results from different versions are not directly comparable.
Coding agent performance
Artificial Analysis Coding Agent Index: per-task performance, API cost and runtime when this model runs inside different agents (Claude Code, Codex, …). Each configuration is evaluated separately.
💳 Subscription Plans
Access DeepSeek V4.1 Flash (1017) through these plans, grouped by product line.
⌨️ Coding plans
2 plansEditor and CLI coding agents, metered in prompts rather than tokens.
Which plan is right for you?
Pro: Ideal for individual professionals with regular usage needs
| Channel | Type | Input / 1M | Output / 1M | Rate Limit | China | vs Official | |
|---|---|---|---|---|---|---|---|
OpenCode Zen💰 Best Price | 聚合平台 | $0.00 | $0.00 | - | - | -100.0% | |
OpenCode Go2 variants Quota: USD | 聚合平台 | $0.15 | $0.60 | - | - | -50.0% | |
Show all price variants for this channelOff-peakDefault $0.15 / $0.60 per 1M time_window Peak $0.30 / $1.20 per 1M time_window | |||||||
Together AI | 聚合平台 | $0.30 | $1.20 | - | - | - | |
Fireworks AI | 聚合平台 | $0.30 | $1.20 | - | - | - | |
DeepSeek2 variants🏢 Official Peak / Off-peak List = peak; off-peak = 50% of peak. Peak: Input $0.30 · cache hit $0.006 · output $1.20 / 1M tokens Off-peak: Input $0.15 · cache hit $0.003 · output $0.60 / 1M tokens | 官方 | $0.30 | $1.20 | - | Baseline | ||
Show all price variants for this channelPeakDefault $0.30 / $1.20 per 1M time_window Off-peak $0.15 / $0.60 per 1M time_window | |||||||
Vercel AI Gateway2 variants | 聚合平台 | $0.30 | $1.20 | - | - | - | |
| [needs-verification] Gateway model id deepseek/deepseek-v4.1-flash-1017. Rates from AI Gateway /v1/models regional.us fields (per-token × 1e6) on 2026-10-18: input $0.30 / output $1.20 / cache-read $0.007 per 1M. Gateway marks varies_by_provider=true and publishes peak_pricing.multiplier=2 with DeepSeek-style UTC peak windows; headline uses regional.us peak-equivalent rates. Cache-read $0.007/1M is Gateway-published (DeepSeek official peak cache-read is $0.006/1M). Changelog claims no markup / at provider pricing but does not list $/1M. Fast variant deepseek/deepseek-v4.1-flash-1017-fast: $0.60 / $2.40 / $0.014 per 1M (Gateway API). Sources: https://ai-gateway.vercel.sh/v1/models ; https://vercel.com/changelog/deepseek-v4-1-flash-1017-now-available-on-ai-gateway | |||||||
Show all price variants for this channelUS regional (Gateway)Default $0.30 / $1.20 per 1M region Fast $0.60 / $2.40 per 1M model_alias | |||||||
| Usage Level | Tokens/Month | OpenCode Zen | OpenCode Go | Together AI | Fireworks AI |
|---|---|---|---|---|---|
| 🐣 Light | 0.1M in + 0.1M out | $0.00 | $0.05 | $0.09 | $0.09 |
| 📊 Medium | 1.0M in + 0.5M out | $0.00 | $0.45 | $0.90 | $0.90 |
| 🚀 Heavy | 10.0M in + 5.0M out | $0.00 | $4.50 | $9.00 | $9.00 |
| 🏢 Enterprise | 100.0M in + 50.0M out | $0.00 | $45.00 | $90.00 | $90.00 |
Frequently asked questions about DeepSeek V4.1 Flash (1017)
How much does the DeepSeek V4.1 Flash (1017) API cost?
The lowest input rate is USD 0 per 1M tokens; output on that channel is USD 0 per 1M tokens.
Which channel has the cheapest DeepSeek V4.1 Flash (1017) API?
OpenCode Zen has the lowest input rate: USD 0 per 1M tokens. Cross-currency ranking uses exchange rates; compare output and cache prices separately.
What is the DeepSeek V4.1 Flash (1017) cache-read price?
OpenCode Zen: USD 0; OpenCode Go: USD 0.003; Fireworks AI: USD 0.006; DeepSeek: USD 0.006; Vercel AI Gateway: USD 0.007 per 1M cache-read tokens.
When was DeepSeek V4.1 Flash (1017) pricing last updated?
Updated on 2026-10-25. This is the latest recorded price update or verification; see the table for individual channel dates.