DeepSeek V4 Flash running in a Trusted Execution Environment (TEE). Hardware attestation evidence is available for independent verification of enclave identity and configuration.
TEE|1M context|$0.30/M input|$0.70/M output
Tinfoil
inference provider · 6 models
Access 6 models served through Tinfoil on AnonRouter's privacy-first gateway, including DeepSeek V4 Flash, Google Gemma 4 31B Instruct, and Llama 3.3 70B. Tinfoil serves every route inside a verified SEV-SNP and NVIDIA CC enclave with remote attestation, so prompts and responses are never exposed to the host operator.
Models
6
Modalities
2
Text, Embeddings
From (input)
$0.05
per 1M tokens
Max context
1M
Private routes
6 / 6
not anonymous-only
Catalog by modality
6 routesText5Embeddings1
Tinfoil models6
Gemma 4 31B is a dense model from Google DeepMind with 31B parameters, delivering frontier-level reasoning performance. It handles text, image, and video input, supports 256K context, function calling, and configurable thinking modes.
TEE|262K context|$0.40/M input|$1.00/M output
Llama 3.3 70B is a text model.
TEE|131K context|$1.75/M input|$2.75/M output
Kimi K3 is an ultra-large-scale, open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at navigating large repositories, using tools, debugging, and iterating against images, logs, tests, and runtime feedback.
TEE|262K context|$4.00/M input|$20.00/M output
GPT OSS 120B running in a Trusted Execution Environment (TEE). OpenAI's open-weight 117B-parameter MoE model with configurable reasoning depth and native tool use, with hardware attestation evidence available for independent verification.
TEE|131K context|$0.15/M input|$0.60/M output
Nomic Embed Text served in a Tinfoil verified confidential enclave.
TEE|8K context|$0.05/M input|Verify/M output