FIND A MODEL FOR YOUR TASK

Appunvs Models

Compare availability, context, reasoning, and RMB pricing. Choose a model for your task, with unconfirmed information clearly marked.

Catalog updated 2026-09-21 Showing the build-time snapshot

Models: 32
Vendors: 9
Active: 30
Last updated: 2026-09-21

Search models

Vendor region

Status

Access

32 models

Mainland

DeepSeek

Vendor docs

DeepSeek V4 Pro

deepseek-v4-pro

Active

Context: 1M
Input: Off-peak ¥4.5 · Peak ¥9 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Output: Off-peak ¥13.5 · Peak ¥27 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Access: Hosted Bring your key

Reasoning request_param · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 384k
Deployment: standard · public · current
Cache read: Off-peak ¥0.15 · Peak ¥0.3 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · responses: supported · fim: supported
Last verified: 2026-08-26
Official sources: models_api pricing thinking_guide chat_api responses_api fim_api kv_cache

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

DeepSeek Flash

deepseek-flash

Active

Context: 1M
Input: Off-peak ¥1 · Peak ¥2 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Output: Off-peak ¥4 · Peak ¥8 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Access: Hosted Bring your key

Reasoning request_param · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 384k
Deployment: standard · public · current
Cache read: Off-peak ¥0.02 · Peak ¥0.04 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Input modalities: text · image
Output modalities: text
Endpoints: chat_completions: supported · responses: supported · fim: supported
Last verified: 2026-09-21
Official sources: models_api pricing thinking_guide chat_api responses_api fim_api kv_cache vision

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

DeepSeek V4 Flash

deepseek-v4-flash

Retired

Context: 1M
Input: Off-peak ¥1.5 · Peak ¥3 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Output: Off-peak ¥4.5 · Peak ¥9 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Access: Unavailable

Reasoning request_param · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 384k
Deployment: standard · public · current
Cache read: Off-peak ¥0.05 · Peak ¥0.1 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · responses: supported · fim: supported
Last verified: 2026-09-21
Official sources: models_api pricing thinking_guide chat_api responses_api fim_api kv_cache

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

DeepSeek V4 Flash Vision Experimental

deepseek-v4-flash-vision-exp

Retired

Context: 1M
Input: Off-peak ¥1.5 · Peak ¥3 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Output: Off-peak ¥4.5 · Peak ¥9 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Access: Unavailable

Reasoning request_param · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 384k
Deployment: standard · public · current
Cache read: Off-peak ¥0.05 · Peak ¥0.1 RMB / 1M tokens Peak: Mon–Fri 09:00–12:00, 14:00–18:00 (Beijing time)
Input modalities: text · image
Output modalities: text
Endpoints: chat_completions: supported · responses: supported
Last verified: 2026-09-21
Official sources: models_api pricing thinking_guide chat_api responses_api kv_cache vision

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Mainland

Moonshot (Kimi)

Vendor docs

Kimi K3

kimi-k3

Active

Context: 1M
Input: ¥20 RMB / 1M tokens
Output: ¥100 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on · max / high / low

View supplementary specs

Max output per request: 1M
Deployment: standard · public · current
Cache read: ¥2 RMB / 1M tokens
Input modalities: text · image · video
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: overview models models_api chat_api parameters reasoning_effort thinking cache structured_output pricing_k3

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Kimi K2.7 Code

kimi-k2.7-code

Active

Context: 262k
Input: ¥6.5 RMB / 1M tokens
Output: ¥27 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Deployment: standard · public · current
Cache read: ¥1.3 RMB / 1M tokens
Input modalities: text · image · video
Output modalities: text
Endpoints: chat_completions: supported · batch: unknown
Last verified: 2026-08-26
Official sources: models models_api chat_api parameters thinking cache pricing_k27 batch_pricing batch_guide
Documentation conflicts: endpoints.batch.model_support

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Kimi K2.7 Code HighSpeed

kimi-k2.7-code-highspeed

Active

Context: 262k
Input: ¥13 RMB / 1M tokens
Output: ¥54 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Deployment: highspeed · base: kimi-k2.7-code · public · current · ≈180 TPS
Cache read: ¥2.6 RMB / 1M tokens
Input modalities: text · image · video
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: models models_api chat_api parameters thinking cache pricing_k27

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Kimi K2.6

kimi-k2.6

Active

Context: 262k
Input: ¥6.5 RMB / 1M tokens
Output: ¥27 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Deployment: standard · public · current
Cache read: ¥1.1 RMB / 1M tokens
Input modalities: text · image · video
Output modalities: text
Endpoints: chat_completions: supported · batch: supported
Last verified: 2026-08-26
Official sources: models models_api chat_api parameters thinking cache pricing_k26 batch_pricing batch_guide official_tools tool_pricing
Documentation conflicts: pricing.extras.web_search

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Mainland

智谱 GLM

Vendor docs

智谱 GLM-5.3

glm-5.3

Active

Context: 1M
Input: Unverified RMB / 1M tokens
Output: Unverified RMB / 1M tokens
Access: Bring your key

Reasoning reasoning_effort · always on · max / high / low

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · responses: supported · anthropic_messages: supported
Last verified: 2026-08-26
Official sources: overview models_api chat_api cache pricing glm53 migration releases responses_api anthropic_api

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

智谱 GLM-5.2

glm-5.2

Active

Context: 1M
Input: ¥8 RMB / 1M tokens
Output: ¥28 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · max / xhigh / high / medium / low / minimal / none

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Cache read: ¥2 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: overview models_api chat_api thinking cache pricing glm52 migration releases

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

智谱 GLM-5.1

glm-5.1

Active

Context: 200k
Input: ¥6 RMB / 1M tokens >32k ¥8
Output: ¥24 RMB / 1M tokens >32k ¥28
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Cache read: ¥1.3 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: overview models_api chat_api thinking cache pricing glm51

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

智谱 GLM-5-Turbo

glm-5-turbo

Active

Context: 200k
Input: ¥5 RMB / 1M tokens >32k ¥7
Output: ¥22 RMB / 1M tokens >32k ¥26
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Cache read: ¥1.2 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: overview models_api chat_api thinking cache pricing glm5_turbo

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

智谱 GLM-4.7

glm-4.7

Active

Context: 200k
Input: ¥2 RMB / 1M tokens >32k ¥4
Output: ¥8 RMB / 1M tokens >32k ¥16
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Max output per request: 131k
Deployment: standard · public · legacy
Cache read: ¥0.4 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: overview models_api chat_api thinking cache pricing glm47 migration

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Mainland

MiniMax (海螺)

Vendor docs

MiniMax M3

MiniMax-M3

Active

Context: 1M
Input: ¥2.1 RMB / 1M tokens >512k ¥4.2
Output: ¥8.4 RMB / 1M tokens >512k ¥16.8
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Max output per request: 524k
Deployment: standard · public · current
Cache read: ¥0.42 RMB / 1M tokens
Input modalities: text · image · video
Output modalities: text
Endpoints: chat_completions: supported · anthropic_messages: supported
Last verified: 2026-08-26
Official sources: overview text_generation models_api openai_api anthropic_api m3_tools cache pricing chat_schema
Documentation conflicts: endpoints.anthropic_messages.models

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

MiniMax M2.7

MiniMax-M2.7

Active

Context: 205k
Input: ¥2.1 RMB / 1M tokens
Output: ¥8.4 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Deployment: standard · public · current · ≈60 TPS
Cache read: ¥0.42 RMB / 1M tokens
Cache write: ¥2.625 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · anthropic_messages: supported
Last verified: 2026-08-26
Official sources: overview text_generation models_api openai_api anthropic_api cache pricing

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

MiniMax M2.7 HighSpeed

MiniMax-M2.7-highspeed

Active

Context: 205k
Input: ¥4.2 RMB / 1M tokens
Output: ¥16.8 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Deployment: highspeed · base: MiniMax-M2.7 · public · current · ≈100 TPS
Cache read: ¥0.42 RMB / 1M tokens
Cache write: ¥2.625 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · anthropic_messages: supported
Last verified: 2026-08-26
Official sources: overview text_generation models_api openai_api anthropic_api cache pricing

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

MiniMax M2

MiniMax-M2

Active

Context: 205k
Input: ¥2.1 RMB / 1M tokens
Output: ¥8.4 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param · always on

View supplementary specs

Deployment: standard · public · legacy
Cache read: ¥0.21 RMB / 1M tokens
Cache write: ¥2.625 RMB / 1M tokens
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported · anthropic_messages: supported
Last verified: 2026-08-26
Official sources: overview text_generation models_api openai_api anthropic_api cache pricing

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Mainland

小米 MiMo

Vendor docs

小米 MiMo V2.5 Pro

mimo-v2.5-pro

Active

Context: 1M
Input: ¥3 RMB / 1M tokens
Output: ¥6 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Cache read: ¥0.025 RMB / 1M tokens
Non-token charges: web_search: ¥16 / 1000 call
Input modalities: text
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: models models_api thinking structured_output web_search pricing

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

小米 MiMo V2.5

mimo-v2.5

Active

Context: 1M
Input: ¥1 RMB / 1M tokens
Output: ¥2 RMB / 1M tokens
Access: Hosted Bring your key

Reasoning request_param

View supplementary specs

Max output per request: 131k
Deployment: standard · public · current
Cache read: ¥0.02 RMB / 1M tokens
Non-token charges: web_search: ¥16 / 1000 call
Input modalities: text · image · audio · video
Output modalities: text
Endpoints: chat_completions: supported
Last verified: 2026-08-26
Official sources: models models_api thinking structured_output web_search pricing

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Global

OpenAI

Vendor docs

GPT-5.5

gpt-5.5

Active

Context: 1.1M
Input: ¥36 RMB / 1M tokens >272k ¥72
Output: ¥216 RMB / 1M tokens >272k ¥324
Access: Bring your key

Reasoning reasoning_effort · xhigh / high / medium / low / none

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥3.6 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: responses: supported · chat_completions: supported · batch: supported
Training knowledge cutoff: 2025-12-01 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models gpt55 model_guidance responses_api chat_api reasoning prompt_cache pricing batch flex fast_mode deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

GPT-5.6 Sol

gpt-5.6-sol

Active

Context: 1.1M
Input: ¥28.8 RMB / 1M tokens >272k ¥57.6
Output: ¥144 RMB / 1M tokens >272k ¥216
Access: Bring your key

Reasoning reasoning_effort · max / xhigh / high / medium / low / none · pro / standard

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥2.88 RMB / 1M tokens
Cache write: ¥36 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: responses: supported · chat_completions: supported · batch: supported
Training knowledge cutoff: 2026-02-16 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models gpt56_sol model_guidance responses_api chat_api reasoning prompt_cache pricing batch flex fast_mode deprecations
Documentation conflicts: pricing.standard

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

GPT-5.6 Terra

gpt-5.6-terra

Active

Context: 1.1M
Input: ¥14.4 RMB / 1M tokens >272k ¥28.8
Output: ¥86.4 RMB / 1M tokens >272k ¥129.6
Access: Bring your key

Reasoning reasoning_effort · max / xhigh / high / medium / low / none · pro / standard

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥1.44 RMB / 1M tokens
Cache write: ¥18 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: responses: supported · chat_completions: supported · batch: supported
Training knowledge cutoff: 2026-02-16 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models gpt56_terra model_guidance responses_api chat_api reasoning prompt_cache pricing batch flex fast_mode deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

GPT-5.6 Luna

gpt-5.6-luna

Active

Context: 1.1M
Input: ¥1.44 RMB / 1M tokens >272k ¥2.88
Output: ¥8.64 RMB / 1M tokens >272k ¥12.96
Access: Bring your key

Reasoning reasoning_effort · max / xhigh / high / medium / low / none · pro / standard

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥0.144 RMB / 1M tokens
Cache write: ¥1.8 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: responses: supported · chat_completions: supported · batch: supported
Training knowledge cutoff: 2026-02-16 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models gpt56_luna model_guidance responses_api chat_api reasoning prompt_cache pricing batch flex fast_mode deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Global

Anthropic Claude

Vendor docs

Claude Opus 5

claude-opus-5

Active

Context: 1M
Input: ¥36 RMB / 1M tokens
Output: ¥180 RMB / 1M tokens
Access: Bring your key

Reasoning adaptive · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥3.6 RMB / 1M tokens
Cache write: ¥45 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: messages: supported · message_batches: supported
Training knowledge cutoff: 2026-05 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models model_ids messages_api thinking extended_thinking effort pricing prompt_cache batch fast_mode deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Claude Sonnet 5

claude-sonnet-5

Active

Context: 1M
Input: ¥14.4 RMB / 1M tokens
Output: ¥72 RMB / 1M tokens
Access: Bring your key

Reasoning adaptive · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥1.44 RMB / 1M tokens
Cache write: ¥18 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: messages: supported · message_batches: supported
Training knowledge cutoff: 2026-01 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models model_ids messages_api thinking extended_thinking effort pricing prompt_cache batch deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Claude Fable 5

claude-fable-5

Active

Context: 1M
Input: ¥72 RMB / 1M tokens
Output: ¥360 RMB / 1M tokens
Access: Bring your key

Reasoning adaptive · always on · max / xhigh / high / medium / low

View supplementary specs

Max output per request: 128k
Deployment: standard · public · current
Cache read: ¥7.2 RMB / 1M tokens
Cache write: ¥90 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: messages: supported · message_batches: supported
Training knowledge cutoff: 2026-01 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models model_ids messages_api thinking extended_thinking effort pricing prompt_cache batch deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Claude Haiku 4.5

claude-haiku-4-5-20251001

Active

Context: 200k
Input: ¥7.2 RMB / 1M tokens
Output: ¥36 RMB / 1M tokens
Access: Bring your key

Reasoning extended_thinking · budget ≥1024

View supplementary specs

Max output per request: 64k
Deployment: standard · public · current
Cache read: ¥0.72 RMB / 1M tokens
Cache write: ¥9 RMB / 1M tokens
Input modalities: text · image
Output modalities: text
Endpoints: messages: supported · message_batches: supported
Training knowledge cutoff: 2025-02 Latest date covered by the model’s training data
Last verified: 2026-08-26
Official sources: models_api models model_ids messages_api thinking extended_thinking pricing prompt_cache batch deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Global

Google Gemini

Vendor docs

Gemini 3.1 Pro (preview)

gemini-3.1-pro-preview

Active

Context: 1M
Input: ¥14.4 RMB / 1M tokens >200k ¥28.8
Output: ¥86.4 RMB / 1M tokens >200k ¥129.6
Access: Bring your key

Reasoning thinking_config · always on · high / medium / low

View supplementary specs

Max output per request: 66k
Deployment: standard · public · current
Cache read: ¥1.44 RMB / 1M tokens
Non-token charges: context_cache_storage: ¥32.4 / 1000000 token_hour · context_cache_storage: ¥58.32 / 1000000 token_hour · google_search_grounding: ¥100.8 / 1000 query · google_maps_grounding: ¥100.8 / 1000 query
Input modalities: text · image · audio · video · pdf
Output modalities: text
Endpoints: interactions: supported · generate_content: supported · batch: supported
Last verified: 2026-08-26
Official sources: models_api models gemini31_pro interactions generate_content thinking pricing caching batch flex priority deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Gemini 3.7 Flash

gemini-3.7-flash

Active

Context: 1M
Input: ¥5.4 RMB / 1M tokens
Output: ¥27 RMB / 1M tokens
Access: Bring your key

Reasoning thinking_config · always on · high / medium / low

View supplementary specs

Max output per request: 66k
Deployment: standard · public · current
Cache read: ¥0.54 RMB / 1M tokens
Non-token charges: context_cache_storage: ¥3.6 / 1000000 token_hour · context_cache_storage: ¥7.2 / 1000000 token_hour · google_search_grounding: ¥100.8 / 1000 query · google_maps_grounding: ¥100.8 / 1000 query
Input modalities: text · image · audio · video · pdf
Output modalities: text
Endpoints: interactions: supported · generate_content: supported · batch: supported
Last verified: 2026-08-26
Official sources: models_api models gemini37_flash interactions generate_content thinking pricing caching batch flex priority deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Gemini 3.6 Flash

gemini-3.6-flash

Active

Context: 1M
Input: ¥5.4 RMB / 1M tokens
Output: ¥27 RMB / 1M tokens
Access: Bring your key

Reasoning thinking_config · always on · high / medium / low / minimal

View supplementary specs

Max output per request: 66k
Deployment: standard · public · current
Cache read: ¥0.54 RMB / 1M tokens
Non-token charges: context_cache_storage: ¥3.6 / 1000000 token_hour · context_cache_storage: ¥7.2 / 1000000 token_hour · google_search_grounding: ¥100.8 / 1000 query · google_maps_grounding: ¥100.8 / 1000 query
Input modalities: text · image · audio · video · pdf
Output modalities: text
Endpoints: interactions: supported · generate_content: supported · batch: supported
Last verified: 2026-08-26
Official sources: models_api models gemini36_flash interactions generate_content thinking pricing caching batch flex priority deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Gemini 3.5 Flash-Lite

gemini-3.5-flash-lite

Active

Context: 1M
Input: ¥2.16 RMB / 1M tokens
Output: ¥18 RMB / 1M tokens
Access: Bring your key

Reasoning thinking_config · always on · high / medium / low / minimal

View supplementary specs

Max output per request: 66k
Deployment: standard · public · current
Cache read: ¥0.216 RMB / 1M tokens
Non-token charges: context_cache_storage: ¥7.2 / 1000000 token_hour · google_search_grounding: ¥100.8 / 1000 query · google_maps_grounding: ¥100.8 / 1000 query
Input modalities: text · image · audio · video · pdf
Output modalities: text
Endpoints: interactions: supported · generate_content: supported · batch: supported
Last verified: 2026-08-26
Official sources: models_api models gemini35_flash_lite interactions generate_content thinking pricing caching batch flex priority deprecations

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Global

xAI Grok

Vendor docs

Grok 4.6

grok-4.6

Active

Context: 500k
Input: ¥14.4 RMB / 1M tokens >200k ¥28.8
Output: ¥43.2 RMB / 1M tokens >200k ¥86.4
Access: Bring your key

Reasoning reasoning_effort · always on · xhigh / high / medium / low

View supplementary specs

Deployment: standard · public · current
Cache read: ¥3.6 RMB / 1M tokens
Non-token charges: web_search: ¥36 / 1000 call · x_search: ¥36 / 1000 call · code_execution: ¥36 / 1000 call
Input modalities: text · image
Output modalities: text
Endpoints: chat_completions: supported · responses: supported
Last verified: 2026-08-26
Official sources: models_api language_models_api models grok46 chat_api responses_api reasoning streaming prompt_cache pricing release_notes

Supplementary specifications come from official sources. Unconfirmed or conflicting information is clearly marked.

Appunvs Models

No matching models