MODEL PROOF
DeepSeek V4 Flash 0423
FRONTIER
Nothing scores at least as high for less, or higher for the same price, under this workload.
Price under confirmation since 2026-10-12. Upstream lists another price for this model now, or other prices from its sellers. What is shown is the last we published; the new figures are published once a second day shows the same values. · Serving precision differs between offers.
- Independent LMArena score
- 1,432.2 ± 4.02 · 52,016 votes
- Context window
- 1.05M
- Maximum output
- 943.72K
- Input modalities
- text
- Output modalities
- text
- Published input price
- $0.090 / 1M tokens
- Published output price
- $0.180 / 1M tokens
- Pricing kind
- fixed
Inspect complete billing conditions and endpoint terms below.
Price from DeepInfra: the cheapest offer that is serving, at standard delivery, with at least 90% of the reference context and at a declared precision that is not 4-bit (fp8), at its standard rate. The maker’s own precision is not known.
Price under confirmation since 2026-10-12 Upstream lists another price for this model now, or other prices from its sellers. What is shown is the last we published; the new figures are published once a second day shows the same values.
Cheaper right now: $0.053/M at StreamLake — promotion 70% It does not pass the like-for-like test, so it never sets a rank.
Every price dimension · USD per million tokens
| Input | $0.090/M |
|---|---|
| Output | $0.180/M |
| Cached input | $0.018/M |
| Cache write | Unknown/M |
| Cache write 1h | Unknown/M |
| Reasoning | Unknown/M |
Who sells it
The same weights, different shops. Cheapest is not like-for-like when serving precision differs.
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.090/M from DeepInfra.
16 provider offers · rates, limits and conditions
Rates are USD per million tokens. Endpoint terms can differ even when the quoted price and precision match. A listed parameter is a provider declaration, not a task-success test.
| Seller | Input | Output | Precision | Uptime (1d) |
|---|---|---|---|---|
| Relacerelace/fp4 Endpoint terms
| $0.0009 | $1.28 | fp4 | 99.91% |
| OpenInferenceopen-inference/fp4 Endpoint terms
| $0.0010 | $1.41 | fp4 | 99.20% |
| StreamLakestreamlake/fp8 Endpoint terms
| $0.042 | $0.084 | fp8 | 99.38% |
| DeepInfradeepinfra/fp8 Endpoint terms
| $0.090 | $0.180 | fp8 | 99.62% |
| GMICloudgmicloud/fp8 Endpoint terms
| $0.091 | $0.182 | fp8 | 99.93% |
| Venicevenice Endpoint terms
| $0.097 | $0.193 | undeclared | 95.11% |
| DigitalOceandigitalocean Endpoint terms
| $0.098 | $0.196 | undeclared | 99.77% |
| SiliconFlowsiliconflow/fp8 Endpoint terms
| $0.130 | $0.280 | fp8 | 99.80% |
| Alibabaalibaba/fp8 Endpoint terms
| $0.134 | $0.268 | fp8 | 99.28% |
| Baidubaidu/fp8 Endpoint terms
| $0.140 | $0.280 | fp8 | 99.06% |
| Novitanovita/fp8 Endpoint terms
| $0.140 | $0.280 | fp8 | 99.85% |
| Parasailparasail/fp8 Endpoint terms
| $0.140 | $0.280 | fp8 | 99.73% |
| AtlasCloudatlas-cloud/fp4 Endpoint terms
| $0.140 | $0.280 | fp4 | 99.36% |
| Mancer 2mancer/fp8 Endpoint terms
| $0.190 | $0.500 | fp8 | 94.89% |
| Azureazure/us Endpoint terms
| $0.210 | $0.560 | undeclared | 95.83% |
| Cloudflarecloudflare Endpoint terms
| $0.440 | $1.32 | undeclared | 98.01% |
Independent scores
| LMArena Elo | 1432.2default |
|---|---|
| LMArena row | deepseek-v4-flash |
LMArena Elo under CC BY 4.0. Absence of another board is not a score of zero.
Best rankings by task
- SVG #41 1155
- Game development #53 1200
- 3D #54 1189
- Websites #57 1209
- Code categories #57 1207
- ASCII art #57 1113
- UI components #71 1164
- Data visualisation #89 1129
Per-task Elo and rank from Design Arena, via OpenRouter’s model feed. The rank is the feed’s, and it counts models OpenRouter does not list.
Capability
| Context window | 1M |
|---|---|
| Max output | 943K |
| Input modes | text |
| Tool use | yes |
| Reasoning | optional |
| Open weights | yes |
Provenance
| Price source | Published record of 2026-10-06, under confirmation since 2026-10-12 |
|---|---|
| Fetched | 2026-10-16 |
| Held | Price and provider offers kept at the published record of 2026-10-06 until a second day confirms the newer upstream value. |
| Quality data | verified |
| Cross-checked | vendor page |
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
Questions this page answers
What does DeepSeek V4 Flash 0423 cost?
$0.090/M in, $0.180/M out: the published record of 2026-10-06, under confirmation since 2026-10-12. That is the API row, not a subscription.
Does DeepSeek V4 Flash 0423 have an independent quality score?
DeepSeek V4 Flash 0423 has an LMArena score in this catalogue.
What beats DeepSeek V4 Flash 0423?
Nothing scores at least as high as DeepSeek V4 Flash 0423 for less, or higher for the same price.
Does this page use Artificial Analysis scores for DeepSeek V4 Flash 0423?
No. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo where a score exists; otherwise the row is unrated.
Is the cheapest DeepSeek V4 Flash 0423 endpoint the same product?
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.090/M from DeepInfra.
MODEL MONUMENT
Save or share this proof
Badge
The verdict as one SVG, for a README or a docs page. It states this model's dominance status at the balanced workload on the LMArena lens, with the date it was computed, and it is rebuilt with the catalogue — so it changes when the verdict changes, including to one you would rather it did not.
[](https://undominated.ai/models/deepseek__deepseek-v4-flash/)<a href="https://undominated.ai/models/deepseek__deepseek-v4-flash/"><img src="https://undominated.ai/badge/deepseek__deepseek-v4-flash.svg" alt="DeepSeek V4 Flash 0423 — Undominated.ai dominance verdict" height="36"></a>Direct file: https://undominated.ai/badge/deepseek__deepseek-v4-flash.svg — a static SVG, written by scripts/build-badges.mjs on every build, so the copy you embed is never
older than the last deploy.
