Gemma 4 31B Abliterated
google/gemma-4-31b-it-abliterated-v2
huihui-ai's second abliteration of Google Gemma 4 31B instruct: refusals removed from the weights, vision and tool calling kept. Answers directly, without a thinking phase. Huihui-ai's v2 checkpoint is a separate modification of Google's Gemma 4 31B instruction model. Its card states that the first five layers are retained without abliteration. Unlid currently disables the thinking phase on this endpoint because its upstream does not separate reasoning reliably from answer text; do not infer active reasoning support from the base model. Unlid declares a 32,768-token context and up to 32,768 output tokens. Input: text, image; output: text. Tool calling is supported. No reasoning mode is declared in the structured catalog. Reasoning tokens, when generated, count toward the output budget and are charged as output. Prices are USD per million tokens from the live catalog; missing cache discounts use the normal input price under Unlid billing rules. Limits and capabilities are provider declarations, not results of an inference test.
Model specifications
- Input
- text, image
- Output
- text
- Context
- 32,768 tokens
- Max output
- 32,768 tokens
- Input price
- $0.144 / 1M tokens
- Output price
- $0.432 / 1M tokens
- Released
- 2026-10-07
Capabilities
- Streaming
- Function calling
- Vision
- JSON mode
Provider pricing, discounts and data privacy
Compare effective provider prices, published discounts, regions, retention policies, training use, compliance, and privacy links by service tier.
Standard service tier
1 available provider · tier input average $0.144 / 1M tokens · tier output average $0.432 / 1M tokens
Unlid
Tier: Standard · Region: US
Pricing
- Input
- $0.144 / 1M tokens
- Output
- $0.432 / 1M tokens
- Cache read
- $0.144 / 1M tokens
No provider discount is currently published.
Data privacy and compliance
- Region
- US
- Zero data retention
- No
- Used for training
- No
- Data collection
- Moderated
- No
- GDPR compliant
- No
- HIPAA compliant
- No
- SOC 2 certified
- No
- BYOK supported
- No
Frequently asked questions
- What is Gemma 4 31B Abliterated?
- huihui-ai's second abliteration of Google Gemma 4 31B instruct: refusals removed from the weights, vision and tool calling kept. Answers directly, without a thinking phase. Huihui-ai's v2 checkpoint is a separate modification of Google's Gemma 4 31B instruction model. Its card states that the first five layers are retained without abliteration. Unlid currently disables the thinking phase on this endpoint because its upstream does not separate reasoning reliably from answer text; do not infer active reasoning support from the base model. Unlid declares a 32,768-token context and up to 32,768 output tokens. Input: text, image; output: text. Tool calling is supported. No reasoning mode is declared in the structured catalog. Reasoning tokens, when generated, count toward the output budget and are charged as output. Prices are USD per million tokens from the live catalog; missing cache discounts use the normal input price under Unlid billing rules. Limits and capabilities are provider declarations, not results of an inference test.
- How much does Gemma 4 31B Abliterated cost?
- Input costs start at $0.144 / 1M tokens and output costs start at $0.432 / 1M tokens. Provider-level prices vary by service tier.
- What is the context length of Gemma 4 31B Abliterated?
- Gemma 4 31B Abliterated supports a 32,768 token context window and up to 32,768 output tokens.
- What capabilities does Gemma 4 31B Abliterated support?
- Gemma 4 31B Abliterated supports Streaming, Function calling, Vision, JSON mode.
- Which providers offer Gemma 4 31B Abliterated?
- Gemma 4 31B Abliterated is available from Unlid.
- How do providers handle data privacy for Gemma 4 31B Abliterated?
- 1 of 1 providers report that customer data is not used for training, and 0 offer zero-data-retention routing. Retention, compliance, and privacy-policy links are listed per provider.