impulse

Inkling

Open weights

Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total.

API id InklingReleased 17 Jul 2026 (Jul 2026)
Input / 1M
$0.95
at DeepInfra
Output / 1M
$4.05
at DeepInfra
Context window
524K
tokens
Max output
524K
tokens per reply
Hosts
5
6 listings
Released
Jul 2026
17 Jul 2026

Pricing

List prices at DeepInfra, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.95$4.05
Cached input
83% off input
$0.16

What it costs a month

At DeepInfra list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$518
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$412
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$719
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$3,252
Use your own numbers

Where to run it

5 hosts serve Inkling. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
DeepInfraInference
Inkling
$0.95 / $4.05
Together AIInference
Inkling
$1.00 / $4.05
+2% vs headline
BasetenInference
inkling
$1.00 / $4.05
+2% vs headline
Fireworks AIInference
inkling
$1.00 / $4.05
+2% vs headline
databricks-inkling
$1.00 / $4.05
+2% vs headline
Show 1 more listings: regions, snapshots and variants
Together AI
Alias
inkling
$1.00 / $4.05
+2% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
TextImages
Produces
Text
Features
Prompt caching
Open weights
License: Apache-2.0
Runs in about 2285.7 GB of GPU memory

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than DeepInfra.

Smaller Independent models

No cheaper Independent model keeps its key capabilities.

Other frontier labs

  • GPT-5.4 Mini
    OpenAI · on OpenAI
    $0.75 / $4.50
    Has every key capability
    2% less
  • Gemini 3.8 Flash
    Google · on Google AI Studio
    $0.75 / $3.75
    Has every key capability
    13% less
  • Grok Build 0.1
    xAI · on xAI
    $1.00 / $2.00
    Has every key capability
    28% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    88% less

Open-weight picks

  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    60% less
  • Inkling Small
    Independent · on DeepInfra
    $0.45 / $1.20
    Has every key capability
    63% less
  • DeepSeek Flash
    DeepSeek · on DeepSeek
    $0.30 / $1.20
    Has every key capability
    70% less
  • Qwen3.7 Plus
    Alibaba Qwen · on Replicate
    $0.276 / $1.10
    Has every key capability
    72% less

Call it

Using DeepInfra's API.

Python
from litellm import completion

response = completion(
    model="deepinfra/Inkling",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Inkling cost?

On DeepInfra, Inkling costs $0.95 per million input tokens and $4.05 per million output tokens. Cached input costs $0.16 per million tokens.

Which APIs offer Inkling?

Inkling runs on 5 APIs: DeepInfra, Together AI, Baseten, Fireworks AI and Databricks.

What is Inkling's context window?

524K tokens, with up to 524K tokens in one reply.

Is Inkling being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Inkling?

Models that cost less and keep most of what it can do: GPT-5.4 Mini on OpenAI (2% less), Gemini 3.8 Flash on Google AI Studio (13% less) and Kimi K2.7 Code on CLF AI Gateway (60% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Inkling released?

Independent released it on 17 Jul 2026.

What is Inkling's API model id?

Use Inkling with DeepInfra. Other hosts use their own ids: inkling on Baseten, inkling on Fireworks AI, databricks-inkling on Databricks.

Sources

Raw catalog entry for Inkling on DeepInfra
{
  "max_output_tokens": 524288,
  "max_input_tokens": 524288,
  "input_cost_per_token": 9.5e-7,
  "output_cost_per_token": 0.00000405,
  "cache_read_input_token_cost": 1.6e-7,
  "supports_prompt_caching": true,
  "litellm_provider": "deepinfra",
  "mode": "chat",
  "supports_function_calling": true,
  "supports_tool_choice": true,
  "supports_response_schema": true,
  "supports_reasoning": true,
  "supports_vision": true,
  "source": "https://deepinfra.com/pricing"
}