Skip to content

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.

  1. Gemma 4 26B A4BGoogleRemove
  2. Gemini 4 HydrogenGoogleRemove
  3. Kimi K3.1 ProMoonShot AIRemove
gemma-4-26b-a4b-it vs gemini-4-hydrogen vs kimi-k3.1-pro
AttributeGemma 4 26B A4Bgemma-4-26b-a4b-itGemini 4 Hydrogengemini-4-hydrogenKimi K3.1 Prokimi-k3.1-pro
Pricing
Input$0.13 / 1M$1.50 / 1M$2.00 / 1M
Output$0.40 / 1M$7.50 / 1M$10.00 / 1M
Cache Write (5m)$0.13 / 1M$1.50 / 1M$2.00 / 1M
Cache Write (1h)$0.13 / 1M$1.50 / 1M$2.00 / 1M
Cache Read$0.13 / 1M$1.50 / 1M$2.00 / 1M
Web Search$0 / 1M$0 / 1M$0 / 1M
Context
Max context262.1K1M1M
Max outputN/AN/AN/A
Capabilities
VisionYesYesYes
Function CallingYesYesYes
JSON ModeYesYesYes
StreamingYesYesYes
Catalogue
ProviderGoogleGoogleMoonShot AI
Categorychatchatchat
Charge typePay As You GoPay As You GoPay As You Go
Released———
Description
SummaryGemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind, featuring 25.2B total parameters with only 3.8B activated per token—delivering near 31B-class quality at a fraction of the compute cost. It supports multimodal inputs including text, images, and video (up to 60s at 1fps). The model includes a 256K token context window, native function calling, configurable thinking/reasoning modes, and structured output support. Released under the Apache 2.0 license, it is well suited for efficient, production-ready multimodal and agentic applications.Gemini 4 Hydrogen is Google's high-efficiency multimodal model in the Gemini 4 family, optimized for high-throughput agentic workflows, coding, and multi-model routing. It delivers strong performance at lower cost and latency than Gemini 4 Carbon, supporting text, image, audio, video, and file inputs for versatile multimodal applications. Designed for scalable production workloads and long-horizon execution, Gemini 4 Hydrogen supports up to 1M output tokens through continuation and configurable reasoning effort levels (low, medium, high, and xhigh) to balance reasoning depth, speed, and cost. It is well suited for coding assistants, autonomous agents, complex workflow automation, and large-scale AI systems.Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model, building on Kimi K3 with enhanced post-training for advanced reasoning and autonomous task execution. It features a 1M-token context window and native image and video understanding, enabling complex, long-context multimodal workflows. Optimized for long-horizon coding, 3D creation, financial research, and end-to-end agentic knowledge work, Kimi K3.1 Pro is well suited for autonomous software engineering, multimodal analysis, and complex professional workflows that require sustained reasoning and reliable execution.