MODEL PROOF
DeepSeek V3.2
BETTER VALUE OPTION
A stored alternative has an equal-or-higher measured score and an equal-or-lower price.
Serving precision differs between offers.
DeepSeek V4.1 Flash · Δ score 38.4 points · 42% lower measured price · $0.227 / 1M tokens
GLM 5.3 Flash is a further stored option with these losses: 147K → 131K max output
- Independent LMArena score
- 1,424.6 ± 3.47 · 47,729 votes
- Context window
- 163.84K
- Maximum output
- 147.46K
- Input modalities
- text
- Output modalities
- text
- Published input price
- $0.259 / 1M tokens
- Published output price
- $0.800 / 1M tokens
- Pricing kind
- fixed
Inspect complete billing conditions and endpoint terms below.
Price from SiliconFlow: 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.
Cheaper right now: $0.290/M at DeepInfra — 4-bit It does not pass the like-for-like test, so it never sets a rank.
Every price dimension · USD per million tokens
| Input | $0.259/M |
|---|---|
| Output | $0.800/M |
| Cached input | $0.140/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 $0.259/M from SiliconFlow.
4 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) |
|---|---|---|---|---|
| SiliconFlowsiliconflow/fp8 Endpoint terms
| $0.259 | $0.800 | fp8 | 99.51% |
| DeepInfradeepinfra/fp4 Endpoint terms
| $0.260 | $0.380 | fp4 | 100.00% |
| Baidubaidu/fp8 Endpoint terms
| $0.280 | $0.420 | fp8 | 100.00% |
| DigitalOceandigitalocean Endpoint terms
| $0.300 | $0.960 | undeclared | 99.17% |
Independent scores
| LMArena Elo | 1424.6default |
|---|---|
| LMArena row | deepseek-v3.2 |
LMArena Elo under CC BY 4.0. Absence of another board is not a score of zero.
Best rankings by task
- ASCII art #71 1068
- Data visualisation #83 1153
- Websites #87 1162
- SVG #89 1016
- UI components #90 1132
- 3D #90 1099
- Code categories #92 1145
- Game development #95 1104
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 | 163K |
|---|---|
| Max output | 147K |
| Input modes | text |
| Tool use | yes |
| Reasoning | optional |
| Open weights | yes |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-11-22 |
| Quality data | verified |
| Cross-checked | vendor page |
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
Questions this page answers
What does DeepSeek V3.2 cost?
$0.259/M in, $0.800/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription.
Does DeepSeek V3.2 have an independent quality score?
DeepSeek V3.2 has an LMArena score in this catalogue.
What beats DeepSeek V3.2?
DeepSeek V4.1 Flash has an equal-or-higher measured score and an equal-or-lower price.
Does this page use Artificial Analysis scores for DeepSeek V3.2?
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 V3.2 endpoint the same product?
Headline cheapest is a lower precision. Like-for-like at the best declared precision is $0.259/M from SiliconFlow.
MODEL MONUMENT