ernie-4.5-vl-28b-a3b-thinking
novita · chat model
Input
$0.3900 / 1M tokens
Output
$0.3900 / 1M tokens
Cached input
N/A
Context
65.5K
Pricing
| Item | Raw value (per token) | Normalized |
|---|---|---|
| Input | 3.9e-7 | $0.3900 / 1M tokens |
| Output | 3.9e-7 | $0.3900 / 1M tokens |
| Embedding | 3.9e-7 | $0.3900 / 1M tokens |
Token limits
Context window
65.5K
Max input tokens
131.1K
Max output tokens
65.5K
Max tokens
65.5K
Capabilities
| Capability | Supported |
|---|---|
| Vision | Yes |
| Function calling | Yes |
| Parallel function calling | Yes |
| Tool choice | Yes |
| Prompt caching | No |
| Reasoning | Yes |
| Response schema | Yes |
| System messages | Yes |
| Audio input | No |
| Audio output | No |
| Web search | No |
| PDF input | No |
| Video input | No |
Similar models
Models with comparable pricing in the same mode (chat).
| Model | Provider | Input | Output | Context | Coding | Features |
|---|---|---|---|---|---|---|
| Meta-Llama-3.1-70B-Instruct | deepinfra | $0.4000 / 1M tokens | $0.4000 / 1M tokens | 131.1K | N/A | Tools |
| Mixtral-8x7B-Instruct-v0.1 | deepinfra | $0.4000 / 1M tokens | $0.4000 / 1M tokens | 32.8K | N/A | Tools |
| DeepSeek-R1 | hyperbolic | $0.4000 / 1M tokens | $0.4000 / 1M tokens | 32.8K | N/A | Tools |
| DeepSeek-V3-0324 | hyperbolic | $0.4000 / 1M tokens | $0.4000 / 1M tokens | 32.8K | N/A | Tools |
| qwen-2.5-72b-instruct | novita | $0.3800 / 1M tokens | $0.4000 / 1M tokens | 8.2K | N/A | Tools |
| granite-ttm-1024-96-r2 | watsonx | $0.3800 / 1M tokens | $0.3800 / 1M tokens | 512 | N/A |
Sources
| Pricing source | LiteLLM model cost map |
| Synced at | 2026-05-27 |
| Manual review | Not reviewed |
Raw LiteLLM fields
{
"litellm_provider": "novita",
"mode": "chat",
"input_cost_per_token": 3.9e-7,
"output_cost_per_token": 3.9e-7,
"max_input_tokens": 131072,
"max_output_tokens": 65536,
"max_tokens": 65536,
"supports_function_calling": true,
"supports_parallel_function_calling": true,
"supports_tool_choice": true,
"supports_vision": true,
"supports_system_messages": true,
"supports_response_schema": true,
"supports_reasoning": true
}