AI Tools Hub
AI Models/Models/DeepSeek V3.2
Back to model explorerlanguage model

DeepSeek V3.2

deepseek/deepseek-v3.2

Model overview, pricing signals, context window, capabilities, and latest tracked AI Gateway endpoint data.

Provider

deepseek

Context window

128K

Max output

8K

Input price

$0.28 / 1M

Model overview
AI Gateway model metadata from the latest Supabase snapshot.

Creator

deepseek

Provider

deepseek

Type

language

Output price

$0.42 / 1M

Tags and capabilities
Used for filtering and future route selection logic.
tool-useimplicit-cachingtool use

Latest route highlights

Best routes in the latest Supabase AI Gateway snapshot. Recommendations are scoped to tracked endpoints only.

Fastest route

556ms

bedrock | deepseek/deepseek-v3.2

Lowest p50 latency in the latest AI Gateway snapshot.

Most reliable

100%

bedrock | deepseek/deepseek-v3.2

Highest recent uptime in the latest AI Gateway snapshot.

Cheapest route

$0.64 / 1M

deepinfra | deepseek/deepseek-v3.2

Lowest combined input plus output price in the latest AI Gateway snapshot.

Balanced route

deepseek

deepseek | deepseek/deepseek-v3.2

Best weighted score in the latest AI Gateway snapshot.

Score 94.6/100. Score = 45% latency rank + 35% uptime rank + 20% price rank across tracked endpoints.

Endpoint comparison
Latest Supabase endpoint snapshot from 5/17/2026, 7:37:40 AM.
Provider routes

5

Latency p50 range

556ms - 1.47s

Best uptime

100%

Snapshot

May 17, 7:37 AM

ProviderStatusLatency p50Latency p95UptimeThroughputInput / 1MOutput / 1MContextMax outputSupported parameters
bedrockbedrock | deepseek/deepseek-v3.2
0556ms957ms100%64 tok/s$0.62 / 1M$1.85 / 1M128K8K
max_tokenstemperaturestoptoolstool_choice
deepseekdeepseek | deepseek/deepseek-v3.2
0652ms803ms100%83 tok/s$0.28 / 1M$0.42 / 1M128K8K
max_tokenstemperaturestoptoolstool_choice
deepinfradeepinfra | deepseek/deepseek-v3.2
0861ms1.4s100%8.5 tok/s$0.26 / 1M$0.38 / 1M163.8K8K
max_tokenstemperaturestoptoolstool_choice
novitanovita | deepseek/deepseek-v3.2
01.47s2.08s100%25 tok/s$0.28 / 1M$0.42 / 1M163.8K65.5K
max_tokenstemperaturestoptoolstool_choicereasoninginclude_reasoning
fireworksfireworks | deepseek/deepseek-v3.2
0————$0.56 / 1M$1.68 / 1M163K163K
max_tokenstemperaturestoptoolstool_choicereasoninginclude_reasoning

Historical trends

Compact trend data generated from Supabase AI Gateway endpoint snapshots. Trends become meaningful after multiple snapshot runs.

Latency p50
Lower is better.
bedrock: 355ms at May 16, 9:58 AMbedrock: 556ms at May 17, 7:24 AMbedrock: 556ms at May 17, 7:37 AMdeepinfra: 633ms at May 16, 9:58 AMdeepinfra: 861ms at May 17, 7:24 AMdeepinfra: 861ms at May 17, 7:37 AMdeepseek: 760ms at May 16, 9:58 AMdeepseek: 652ms at May 17, 7:24 AMdeepseek: 652ms at May 17, 7:37 AMnovita: 1.6s at May 16, 9:58 AMnovita: 1.47s at May 17, 7:24 AMnovita: 1.47s at May 17, 7:37 AM
May 16, 9:58 AMMay 17, 7:37 AM
bedrock
556ms
deepinfra
861ms
deepseek
652ms
novita
1.47s

Updated May 17, 7:37 AM

Uptime
Latest availability samples.
bedrock: 100% at May 16, 9:58 AMbedrock: 100% at May 17, 7:24 AMbedrock: 100% at May 17, 7:37 AMdeepinfra: 100% at May 16, 9:58 AMdeepinfra: 100% at May 17, 7:24 AMdeepinfra: 100% at May 17, 7:37 AMdeepseek: 100% at May 16, 9:58 AMdeepseek: 100% at May 17, 7:24 AMdeepseek: 100% at May 17, 7:37 AMnovita: 100% at May 16, 9:58 AMnovita: 100% at May 17, 7:24 AMnovita: 100% at May 17, 7:37 AM
May 16, 9:58 AMMay 17, 7:37 AM
bedrock
100%
deepinfra
100%
deepseek
100%
novita
100%

Updated May 17, 7:37 AM

Throughput p50
Output token speed by provider route.
bedrock: 95 tok/s at May 16, 9:58 AMbedrock: 64 tok/s at May 17, 7:24 AMbedrock: 64 tok/s at May 17, 7:37 AMdeepinfra: 20 tok/s at May 16, 9:58 AMdeepinfra: 8.5 tok/s at May 17, 7:24 AMdeepinfra: 8.5 tok/s at May 17, 7:37 AMdeepseek: 77.5 tok/s at May 16, 9:58 AMdeepseek: 83 tok/s at May 17, 7:24 AMdeepseek: 83 tok/s at May 17, 7:37 AMnovita: 27 tok/s at May 16, 9:58 AMnovita: 25 tok/s at May 17, 7:24 AMnovita: 25 tok/s at May 17, 7:37 AM
May 16, 9:58 AMMay 17, 7:37 AM
bedrock
64 tok/s
deepinfra
8.5 tok/s
deepseek
83 tok/s
novita
25 tok/s

Updated May 17, 7:37 AM

Price history
Input plus output price per 1M tokens.
bedrock: $2.47 at May 16, 9:58 AMbedrock: $2.47 at May 17, 7:24 AMbedrock: $2.47 at May 17, 7:37 AMdeepinfra: $0.64 at May 16, 9:58 AMdeepinfra: $0.64 at May 17, 7:24 AMdeepinfra: $0.64 at May 17, 7:37 AMdeepseek: $0.70 at May 16, 9:58 AMdeepseek: $0.70 at May 17, 7:24 AMdeepseek: $0.70 at May 17, 7:37 AMfireworks: $2.24 at May 16, 9:58 AMfireworks: $2.24 at May 17, 7:24 AMfireworks: $2.24 at May 17, 7:37 AMnovita: $0.70 at May 16, 9:58 AMnovita: $0.70 at May 17, 7:24 AMnovita: $0.70 at May 17, 7:37 AM
May 16, 9:58 AMMay 17, 7:37 AM
bedrock
$2.47
deepinfra
$0.64
deepseek
$0.70
fireworks
$2.24
novita
$0.70

Updated May 17, 7:37 AM

AI Gateway data status
Supabase snapshots ready

Tracked models

10

23 endpoint rows

Chart models

10

3 snapshot hours

Models updated

May 17, 7:37 AM

Fresh Supabase snapshot

Endpoint updated

May 17, 7:37 AM

Newest snapshot May 17, 7:37 AM

Data note

Model and endpoint data are read from Supabase AI Gateway snapshots generated by the ingestion pipeline. Pricing, routing, and availability may change, so verify with official providers before production use.