New AI Models Added: GPT-5 Codex, Gemini Flash Latest, and Qwen3-Max
Date: 2025-09-28
Summary
We announce the addition of six new AI models from OpenAI, Google, and Alibaba, including the advanced GPT-5 Codex optimized for coding tasks, Gemini Flash Latest with hybrid reasoning capabilities, and Qwen3-Max with enhanced agent programming features. These models expand our platform's capabilities in coding, reasoning, and cost-effective AI solutions.
Details
OpenAI
gpt-5-codex: Advanced coding model optimized for agentic coding tasks with reasoning capabilities and 400K context window. Available in Responses API with regular model snapshot updates.
Key Features:
- Context Window: 400,000 tokens for extensive code analysis and generation
- Advanced Capabilities: Reasoning token support, function calling, structured outputs
- Specialized Focus: Optimized for agentic coding tasks in Codex environments
- Endpoint Support: Available on v1/chat/completions, v1/responses, v1/assistants, v1/batch
Pricing Details:
| Model | Input | Cached Input | Output |
|---|---|---|---|
| gpt-5-codex | $1.25/1M tokens | $0.125/1M tokens | $10.00/1M tokens |
Google
gemini-flash-latest: Points to gemini-2.5-flash-preview-09-2025, our hybrid reasoning model with 1M token context window and thinking budgets.gemini-2.5-flash-preview-09-2025: Latest preview of Gemini 2.5 Flash with enhanced reasoning capabilities and extensive context handling.gemini-flash-lite-latest: Points to gemini-2.5-flash-lite-preview-09-2025, the most cost-effective model built for at-scale usage.gemini-2.5-flash-lite-preview-09-2025: Smallest and most cost-effective model in the Gemini 2.5 series with optimized pricing.
Key Features:
- Context Window: Up to 1M tokens for extensive document processing
- Advanced Capabilities: Hybrid reasoning, thinking budgets, function calling
- Dual Pricing Tiers: Standard Flash and cost-effective Flash-Lite variants
- Latest Aliases: Flash-latest and Flash-lite-latest point to the newest preview versions
Pricing Details:
| Model | Input | Cached Input | Output |
|---|---|---|---|
| gemini-flash-latest | $0.30/1M tokens | $0.15/1M tokens | $2.50/1M tokens |
| gemini-2.5-flash-preview-09-2025 | $0.30/1M tokens | $0.15/1M tokens | $2.50/1M tokens |
| gemini-flash-lite-latest | $0.10/1M tokens | $0.05/1M tokens | $0.40/1M tokens |
| gemini-2.5-flash-lite-preview-09-2025 | $0.10/1M tokens | $0.05/1M tokens | $0.40/1M tokens |
Alibaba
qwen3-max: Latest Qwen 3 series Max model with specialized upgrades in agent programming and tool invocation, achieving state-of-the-art performance for complex agent scenarios.
Key Features:
- Context Window: Advanced context handling with tiered pricing above 32K and 128K tokens
- Advanced Capabilities: Enhanced agent programming, tool invocation, complex reasoning
- Specialized Focus: Optimized for agents operating in complex scenarios
- Performance: State-of-the-art (SOTA) performance in agent programming tasks
Pricing Details:
| Model | Input | Cached Input | Output | Special Pricing |
|---|---|---|---|---|
| qwen3-max | $1.20/1M tokens | $0.10/1M tokens | $6.00/1M tokens | Above 32K: Input $2.40/1M, Output $12.00/1M |
| qwen3-max | - | - | - | Above 128K: Input $3.00/1M, Output $15.00/1M |
Usage Examples
bash
curl https://api.avalai.ir/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $AVALAI_API_KEY" \
-d '{
"model": "gpt-5-codex",
"messages": [
{
"role": "user",
"content": "Write a Python function to implement a binary search tree with insert, search, and delete operations."
}
]
}'python
from openai import OpenAI
client = OpenAI(api_key="your-avalai-api-key", base_url="https://api.avalai.ir/v1")
completion = client.chat.completions.create(
model="gpt-5-codex",
messages=[
{
"role": "user",
"content": "Write a Python function to implement a binary search tree with insert, search, and delete operations.",
}
],
)
print(completion.choices[0].message.content)javascript
import { OpenAI } from "openai";
const client = new OpenAI({
apiKey: process.env.AVALAI_API_KEY,
baseURL: "https://api.avalai.ir/v1",
});
const completion = await client.chat.completions.create({
model: "gpt-5-codex",
messages: [
{
role: "user",
content: "Write a Python function to implement a binary search tree with insert, search, and delete operations.",
},
],
});
console.log(completion.choices[0].message.content);Advanced Features
Reasoning Capabilities
GPT-5 Codex and Gemini Flash models support advanced reasoning with thinking processes visible in responses:
bash
curl https://api.avalai.ir/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $AVALAI_API_KEY" \
-d '{
"model": "gemini-flash-latest",
"messages": [
{
"role": "user",
"content": "Design a scalable microservices architecture for an e-commerce platform"
}
],
"max_tokens": 2048
}'python
response = client.chat.completions.create(
model="gemini-flash-latest",
messages=[
{
"role": "user",
"content": "Design a scalable microservices architecture for an e-commerce platform",
}
],
max_tokens=2048,
)
# The model will show its thinking process before providing the final answer
print(response.choices[0].message.content)javascript
const response = await client.chat.completions.create({
model: "gemini-flash-latest",
messages: [
{
role: "user",
content: "Design a scalable microservices architecture for an e-commerce platform",
},
],
max_tokens: 2048,
});
// The model will show its thinking process before providing the final answer
console.log(response.choices[0].message.content);