Compare models
Put up to 4 models beside each other — token prices, context windows, capabilities and provider, from the same catalogue the model pages read.
| Attribute | Gemini 3.5 Transcribegemini-3.5-transcribe | Gemini 3.1 Flash Image (Google AI Studio)gemini-3.1-flash-image-aistudio | GLM 5.3 Flashglm-5.3-flash |
|---|---|---|---|
| Pricing | |||
| Input | — Not priced per input token | — | $0.075 / 1M |
| Output | — Not priced per output token | — | $0.25 / 1M |
| Cache Write (5m) | Not applicable | Not applicable | $0.075 / 1M |
| Cache Write (1h) | Not applicable | Not applicable | $0.075 / 1M |
| Cache Read | Not applicable | Not applicable | $0.075 / 1M |
| Web Search | $0 / 1M | — | $0 / 1M |
| Request | — | $0.067 / request | — |
| Billing | — | Pay Per Request | — |
| Context | |||
| Max context | 98.3K | N/A | 1M |
| Max output | N/A | N/A | N/A |
| Capabilities | |||
| Vision | Yes | Yes | Yes |
| Function Calling | Yes | No | Yes |
| JSON Mode | No | No | Yes |
| Streaming | No | No | Yes |
| Catalogue | |||
| Provider | Z.AI | ||
| Category | voice | image | chat |
| Charge type | Pay As You Go | Pay Per Request | Pay As You Go |
| Released | 2026-09-25 | — | — |
| Description | |||
| Summary | Google Gemini 3.5 Transcribe speech-to-text. Billed per input and output token. | Gemini 3.1 Flash Image served from Google AI Studio. Billed per image at US$0.067. Call it through /v1/chat/completions with stream set to false; the image is returned in the message content as a base64 data URI. | GLM-5.3-Flash is Z.AI's efficient native multimodal model, designed for coding and long-horizon agentic workflows. It combines strong multimodal capabilities with an architecture optimized for responsive, cost-efficient task execution. Built on a hybrid sparse and linear attention architecture, GLM-5.3-Flash maintains accurate long-context behavior while reducing computational overhead, making it well suited for coding agents, extended multi-step tasks, and scalable production workloads. |