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. Gemini 4 HydrogenGoogleRemove
  2. Gemini 4 CarbonGoogleRemove
  3. GPT OSS 20BOpenAIRemove
gemini-4-hydrogen vs gemini-4-carbon vs gpt-oss-20b
AttributeGemini 4 Hydrogengemini-4-hydrogenGemini 4 Carbongemini-4-carbonGPT OSS 20Bgpt-oss-20b
Pricing
Input$1.50 / 1M$2.00 / 1M$0.075 / 1M
Output$7.50 / 1M$10.00 / 1M$3.00 / 1M
Cache Write (5m)$1.50 / 1M$2.00 / 1M$0.075 / 1M
Cache Write (1h)$1.50 / 1M$2.00 / 1M$0.075 / 1M
Cache Read$1.50 / 1M$2.00 / 1M$0.075 / 1M
Web Search$0 / 1M$0 / 1M$0 / 1M
Context
Max context1M1M131.1K
Max outputN/AN/AN/A
Capabilities
VisionYesYesNo
Function CallingYesYesYes
JSON ModeYesYesYes
StreamingYesYesYes
Catalogue
ProviderGoogleGoogleOpenAI
Categorychatchatchat
Charge typePay As You GoPay As You GoPay As You Go
Released———
Description
SummaryGemini 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.Gemini 4 Carbon is Google's frontier multimodal reasoning model, designed for advanced coding, agentic workflows, and complex multi-step problem solving. It supports text, image, audio, video, and file inputs, enabling integrated reasoning and analysis across diverse modalities. Optimized for long-horizon execution and demanding professional workloads, Gemini 4 Carbon supports up to 1M output tokens through continuation and configurable reasoning effort levels (low, medium, high) to balance reasoning depth, latency, and cost. It is well suited for autonomous agents, software engineering, multimodal analysis, and large-scale workflow automation.gpt-oss-20b is an open-weight, 21B-parameter OpenAI model released under Apache 2.0. It uses a Mixture-of-Experts design so only ~3.6B parameters run each step, enabling faster, lower-cost inference on consumer or single-GPU hardware. Trained in the Harmony format, it supports configurable reasoning depth, fine-tuning, function calling, tool use, and structured outputs.