Models

Anthropic’s latest generally available frontier model for demanding coding and knowledge work. 1M context, 128k output.

Claude Opus 4.7 through AWS Bedrock Global application inference profile.

Claude Opus 4.8 through Amazon Bedrock Converse API. 1M context, 128k output.

Claude Opus 5 through AWS Bedrock Global application inference profiles.

Claude Sonnet 5 through AWS Bedrock Global application inference profiles.

V4 flagship reasoning — 1M context, official direct-API pricing.

文本生成与对话。

Seedance 2.0 文生视频，支持异步生成与下载，按实际视频 tokens 计费。

Mid-tier multimodal. Image + video + text input. 1M context.

Cheapest 1M-context tier. Great for high-volume agents.

Google’s flagship reasoning model + Computer Use. 1M context.

Stable Gemini 3.5 model built for fast frontier intelligence, agents, coding, and grounding.

Google’s current low-cost GA model for high-volume multimodal workloads. 1M context.

Google’s current GA Flash model for fast multimodal reasoning, coding, and agents. 1M context.

Google’s latest GA Flash model for long-horizon coding, autonomous agents, and complex workflows. Introductory pricing through 2026-12-31.

Gemini Omni Flash 1.1 — generally available video generation and editing with audio, 3–10 second output, 360p–4K resolution, and first/last-frame interpolation. Video output is billed separately at approximately $0.10/s.

Backup route for Gemini Omni Flash 1.1 using the legacy Google API credential.

GLM 5 base flagship — released 2026-02-12, available on Pro and Max coding tiers.

Zhipu’s April 2026 flagship — SOTA on SWE-Bench Pro, 200k context, 128k max output.

Z.AI’s current flagship with a 1M-token context window.

文本生成与对话。

OpenAI model via an OpenAI-compatible API.

OpenAI model via an OpenAI-compatible API.

OpenAI model via an OpenAI-compatible API.

OpenAI model via an OpenAI-compatible API.

OpenAI model via an OpenAI-compatible API.

OpenAI model via an OpenAI-compatible API.

OpenAI image gen — flexible sizes, high-fidelity image input.

Flagship multimodal. 1M context. Computer use, MCP, hosted shell.

Efficient GPT-5.6 tier for high-volume workloads. 1.05M context, 128k output.

Previous-generation flagship for complex professional reasoning and coding. 1.05M context, 128k output.

Balanced GPT-5.6 tier with frontier capability at lower cost. 1.05M context, 128k output.

OpenAI’s latest and most capable model for end-to-end reasoning, coding, computer use, and research. Base pricing applies through 272k input tokens; Azure long-context rates apply above that threshold. 1.05M context, 128k output.

根据首帧图片与文字提示生成视频，支持异步生成与下载。

根据参考图片与文字提示生成视频，支持异步生成与下载。

根据文字描述生成视频，支持异步任务与下载。

视频编辑，支持输入视频与文字指令，按输入与输出视频的合计时长计费。

根据首帧图片与文字提示生成视频，支持异步生成与下载。

根据参考图片与文字提示生成视频，支持异步生成与下载。

根据文字描述生成视频，支持异步任务与下载。

Legacy multimodal release. Unavailable to newly registered users and scheduled to sunset on 2026-08-31.

Moonshot multimodal model with 256k context. Official direct-API pricing.

Moonshot’s current coding and agent model with a 256k context window.

Moonshot’s current flagship with text, image and video input and a 1M-token context window.

Prior-gen M-series — same base price, slightly weaker quality.

Original Nano Banana (Gemini 2.5 Flash Image). Fast, fun edits.

Nano Banana 2 (Gemini 3.1 Flash Image) — Pro features at Flash speed. $0.067/1K, $0.10/2K, $0.15/4K.

Nano Banana 2 Lite (Gemini 3.1 Flash-Lite Image) — low-latency 1K image generation and editing.

Nano Banana Pro (Gemini 3 Pro Image) — best fidelity, complex prompts, accurate text. $0.134/1K-2K, $0.24/4K.

根据文字描述生成图片。

根据文字描述生成图片。

文本生成与对话。

文本生成与对话。

Vision-language flagship — strong OCR, document QA, GUI grounding.

文本生成与对话。

文本生成与对话。

文本生成与对话。

文本生成与对话。

Alibaba’s current Qwen flagship with a 1M-token context window. Stored price is the <=128k-input tier; 128k–256k requests cost $2/$12 per MTok.

文本生成与对话。

文本生成与对话。

文本生成与对话。

文本生成与对话。

文本生成与对话。

Cinematic text/image-to-video with native audio at 720p, 1080p, or 4K.

Veo 3.1 Fast — $0.10/s @720p, $0.12/s @1080p, $0.30/s @4K.

根据文字描述生成视频，支持异步任务与下载。