AIbase
EN
Home
AI Models Finder
Sonar Reasoning
S

Sonar Reasoning

新版 Sonar 搜索模型。該模型由 Llama 3.3 70B 進一步訓練而來，對搜索應用進行了優化，運行在 Cerebras 推理基礎設施上。 Perplexity 表示新版 Sonar 的 Token 解碼速度達到 1200 個每秒，是 Gemini 2.0 Flash 的 8.5 倍以上，“幾乎可以即時生成答案”。
Intelligence(Medium)
Speed(Relatively Slow)
Input Supported Modalities
Yes
Is Reasoning Model
127,000
Context Window
0
Maximum Output Tokens
-
Knowledge Cutoff
Go Compare

Pricing

- /M tokens
Input
- /M tokens
Output
￥14.4 /M tokens
Blended Price

Quick Simple Comparison

Input

Output

Sonar Pro
R1 1776
Sonar

Basic Parameters

Sonar ReasoningTechnical Parameters
Parameter Count
Not Announced
Context Length
127.00k tokens
Training Data Cutoff
Open Source Category
Proprietary
Multimodal Support
Text Only
Throughput
0
Release Date
2025-01-28
Response Speed
84.495,995 tokens/s

Benchmark Scores

Below is the performance of Sonar Reasoning in various standard benchmark tests. These tests evaluate the model's capabilities in different tasks and domains.
Intelligence Index
48.83
Large Language Model Intelligence Level
Coding Index
-
Indicator of AI model performance on coding tasks
Math Index
84.57
Capability indicator in solving mathematical problems, mathematical reasoning, or performing math-related tasks
MMLU Pro
-
Massive Multitask Multimodal Understanding - Testing understanding of text, images, audio, and video
GPQA
62.3
Graduate Physics Questions Assessment - Testing advanced physics knowledge with diamond science-level questions
HLE
-
The model's comprehensive average score on the Hugging Face Open LLM Leaderboard
LiveCodeBench
-
Specific evaluation focused on assessing large language models' ability in real-world code writing and solving programming competition problems
SciCode
-
The model's capability in code generation for scientific computing or specific scientific domains
HumanEval
-
Score achieved by the AI model on the specific HumanEval benchmark test set
Math 500 Score
92.1
Score on the first 500 larger, more well-known mathematical benchmark tests
AIME Score
77
An indicator measuring an AI model's ability to solve high-difficulty mathematical competition problems (specifically AIME level)
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase