ToolBunny Logo
ToolBunny
Play Store
Back to Homepage

System Parameter Cheat Sheet

Quick-reference directory mapping system constraints across global AI providers.

GPT-4o

OpenAI · gpt-4o
Context Limit
128,000 tokens
Max Output
16,384 tokens
Input Cost / 1M
$2.50
Output Cost / 1M
$10.00
Max Tokens Param:max_completion_tokens
System Prompt Param:messages[0] (role: 'system')
VisionStructured OutputTools

Flagship multimodal model with native JSON Schema enforcement and audio support.

GPT-4o Mini

OpenAI · gpt-4o-mini
Context Limit
128,000 tokens
Max Output
16,384 tokens
Input Cost / 1M
$0.15
Output Cost / 1M
$0.60
Max Tokens Param:max_completion_tokens
System Prompt Param:messages[0] (role: 'system')
VisionStructured OutputTools

Ultra-fast, cost-efficient model for high-throughput lightweight tasks.

Claude 3.5 Sonnet

Anthropic · claude-3-5-sonnet-20241022
Context Limit
200,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$3.00
Output Cost / 1M
$15.00
Max Tokens Param:max_tokens (required)
System Prompt Param:system (top-level string param)
VisionStructured OutputTools

Top-tier coding, reasoning, and visual layout parsing model.

Claude 3.5 Haiku

Anthropic · claude-3-5-haiku-20241022
Context Limit
200,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$1.00
Output Cost / 1M
$5.00
Max Tokens Param:max_tokens (required)
System Prompt Param:system (top-level string param)
Structured OutputTools

Lightweight text model matching Claude 3 Opus intelligence on benchmarks.

Gemini 1.5 Pro

Google · gemini-1.5-pro
Context Limit
2,000,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$1.25
Output Cost / 1M
$5.00
Max Tokens Param:maxOutputTokens
System Prompt Param:systemInstruction.parts[0].text
VisionStructured OutputTools

Industry-leading 2 million token context window for entire codebase ingestion.

Gemini 1.5 Flash

Google · gemini-1.5-flash
Context Limit
1,000,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$0.075
Output Cost / 1M
$0.30
Max Tokens Param:maxOutputTokens
System Prompt Param:systemInstruction.parts[0].text
VisionStructured OutputTools

Sub-second multimodal latency for high-frequency search and summarization.

DeepSeek R1

DeepSeek · deepseek-reasoner
Context Limit
64,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$0.55
Output Cost / 1M
$2.19
Max Tokens Param:max_tokens
System Prompt Param:messages[0] (role: 'system')
Structured Output

State-of-the-art open reasoning model utilizing chain-of-thought <think> tokens.

Llama 3.3 70B

Meta · meta-llama/llama-3.3-70b-instruct
Context Limit
128,000 tokens
Max Output
4,096 tokens
Input Cost / 1M
$0.50
Output Cost / 1M
$0.75
Max Tokens Param:max_tokens
System Prompt Param:messages[0] (role: 'system')
Structured OutputTools

Open-weights 70B parameter model matching Llama 3.1 405B capabilities.

Mistral Large 2

Mistral · mistral-large-latest
Context Limit
128,000 tokens
Max Output
8,192 tokens
Input Cost / 1M
$2.00
Output Cost / 1M
$6.00
Max Tokens Param:max_tokens
System Prompt Param:messages[0] (role: 'system')
Structured OutputTools

Top European open LLM with strong multilingual & function calling support.

About This Tool

The System Parameter Cheat Sheet is a technical reference matrix designed for AI software engineers, system architects, and prompt engineers building multi-provider LLM applications.

Each major AI provider — OpenAI, Anthropic, Google Cloud Vertex, DeepSeek, Meta, and Mistral — enforces different context window bounds, maximum completion tokens, API parameter key names (e.g. OpenAI max_completion_tokens vs Anthropic max_tokens vs Gemini maxOutputTokens), pricing structures, and system prompt parameters. This cheat sheet provides searchable directory cards and side-by-side matrix comparisons to prevent integration bugs.

How to Use

  1. Filter by Provider or Search: Use the search bar to find specific model names or click provider tags (OpenAI, Anthropic, Google, DeepSeek, Meta, Mistral).
  2. Inspect Technical Directory Cards: Review context limits, max output tokens, 1M token pricing rates, vision capability badges, and exact API parameter keys.
  3. Switch to Side-by-Side Compare: Click + Compare on cards or switch to the Compare view mode to evaluate up to 4 models side-by-side.

Common Use Cases & FAQ

Related Tools

Infrastructure Node Status

Real-time health monitor and latency ping checker for AI API gateways, cloud endpoints, and developer infrastructure nodes.

System Prompt Injector Sandbox

Audit vulnerability structures by mounting mock payloads into structural setups. Test LLM system instructions against prompt injection, jailbreaks, and delimiter attacks.

Password Strength Analyzer

Test the exact strength of your passwords against realistic cracking algorithms. Discover how long it would take a hacker to guess your password.

RSA Key Pair Generator

Generate secure RSA Public and Private Key pairs securely within your browser using the native Web Crypto API. Export to standard PEM format instantly.