NEAR AI Cloud integrates Intel-signed attestation token for
The integration enhances trust and security in AI workloads, streamlining verification processes and potentially setting a new industry standard.
TL;DR
NEAR AI Cloud integrated Anthropic’s Claude Sonnet 5.5 on August 12, 2026, offering a million‑token context window and a 70.6% Terminal‑Bench score at Anthropic’s public pricing, with payment via NEAR token staking. The move blends high‑performance coding AI with blockchain‑native economics, positioning the platform between centralized cloud providers and decentralized inference networks.
Key facts
- NEAR AI Cloud added Claude Sonnet 5.5 on August 12, 2026.
- Pricing is $2 per million input tokens and $10 per million output tokens.
- Inference credits are funded by staking NEAR tokens, which are burned on API calls.
On August 12, 2026, NEAR AI Cloud announced the integration of Anthropic’s Claude Sonnet 5.5 into its blockchain‑native AI platform. The new model, featuring a one‑million‑token context window and adaptive reasoning, is offered through NEAR’s unified gateway and priced at $2 per million input tokens and $10 per million output tokens. The rollout positions NEAR AI Cloud as a middle ground between centralized cloud AI services and decentralized inference networks, leveraging NEAR token staking for payment and verifiable inference.
Background
NEAR AI Cloud, a platform that blends artificial‑intelligence inference with blockchain technology, has been expanding its catalog of frontier models. It operates a unified gateway at cloud‑api.near.ai that supports both OpenAI‑compatible and Anthropic‑native API formats, allowing developers to switch providers without rewriting integration layers. The platform’s architecture traditionally runs models inside Trusted Execution Environments (TEEs) to provide privacy‑preserving, verifiable inference, with payment handled via the NEAR token. Users can stake NEAR to fund compute credits, which are burned against API calls. Prior to this announcement, NEAR AI Cloud had already become a provider on OpenRouter, a service that routes requests across multiple AI providers, and it had positioned itself against centralized AI offerings from AWS Bedrock, Google Cloud, and Azure, as well as decentralized inference networks such as Akash. The addition of Claude Sonnet 5.5, Anthropic’s latest coding‑focused model, follows a broader industry trend of integrating large language models with blockchain‑based payment and verification mechanisms, aiming to combine the reliability of mainstream cloud services with the transparency and token economics of decentralized platforms.
What happened
According to cryptobriefing.com, NEAR AI Cloud added Anthropic’s Claude Sonnet 5.5 to its model lineup on August 12, 2026. The integration is delivered through NEAR’s unified gateway at cloud‑api.near.ai, supporting both OpenAI‑compatible and Anthropic‑native API formats, which allows developers to point existing OpenAI‑style integrations at NEAR’s endpoint without code changes. Claude Sonnet 5.5 runs as an “Incognito” model, meaning it connects via a proxied link to Anthropic’s infrastructure rather than being hosted inside NEAR’s Trusted Execution Environments, resulting in different privacy guarantees that are non‑retentive. The model offers a one‑million‑token context window and adaptive reasoning enabled by default. Pricing mirrors Anthropic’s public rates, set at $2 per million input tokens and $10 per million output tokens. Performance metrics show Claude Sonnet 5.5 achieving a 70.6 % score on the Terminal‑Bench 4.0 coding benchmark, a substantial increase from the 10.3 % score recorded by its predecessor, Sonnet 5. The platform continues to use the NEAR token as the payment rail, with users staking NEAR to fund inference credits that are burned when API calls are made. The announcement also notes that NEAR AI Cloud is now listed on OpenRouter, enabling developers to compare and switch between AI providers through a single routing service.
Market impact
The integration brings a model priced at $2 per million input tokens and $10 per million output tokens to the NEAR AI Cloud marketplace, aligning with Anthropic’s public pricing structure. Claude Sonnet 5.5’s 70.6 % Terminal‑Bench 4.0 score positions it as a high‑performance coding model, potentially attracting developers seeking autonomous coding agents that require long context windows and strong reasoning. By allowing NEAR token staking to fund inference credits, the platform may see increased demand for NEAR token staking, as users convert tokens into compute credits that are subsequently burned. The “Incognito” deployment model differentiates the offering from NEAR’s TEE‑hosted models, possibly influencing user preferences based on privacy and data‑retention considerations. As NEAR AI Cloud competes with centralized providers like AWS Bedrock, Google Cloud, and Azure, as well as decentralized networks such as Akash, the pricing parity and blockchain‑native features could shift a portion of AI inference spend toward token‑based, verifiable services.
Why it matters
The addition of Claude Sonnet 5.5 gives NEAR AI Cloud a coding model that scores 70.6 % on Terminal‑Bench 4.0, a benchmark that measures autonomous coding capabilities. This performance leap suggests the platform can support more sophisticated autonomous coding agents that write, test, and deploy code with minimal human oversight, a capability increasingly sought after in AI‑driven development pipelines. The one‑million‑token context window enables the model to process extensive codebases, enhancing its utility for large‑scale software projects. By offering the model through a blockchain‑native payment system, NEAR AI Cloud merges token economics with AI inference, potentially attracting developers who value verifiable, privacy‑preserving compute and who wish to use NEAR tokens for payment. The “Incognito” deployment model, while differing from NEAR’s TEE‑hosted options, provides non‑retentive interactions, addressing privacy concerns for certain use cases. Overall, the move positions NEAR AI Cloud as a hybrid solution that blends the reliability of mainstream cloud AI services with the transparency and token‑based incentives of decentralized platforms, influencing how developers choose inference providers.
DeFi Intel’s take
Editorial analysis, not a statement of fact.
DeFi Intel's take: The integration of Claude Sonnet 5.5 suggests NEAR AI Cloud is solidifying its role as a bridge between traditional cloud AI services and decentralized inference networks. By matching Anthropic’s public pricing while adding token‑based staking for compute credits, the platform may attract developers looking for cost‑predictable, verifiable AI workloads. The performance jump to a 70.6 % Terminal‑Bench 4.0 score indicates the model could power more autonomous coding agents, which could increase demand for NEAR token staking and burn mechanisms. However, the “Incognito” deployment means the model does not benefit from NEAR’s TEE‑based privacy guarantees, which could steer privacy‑sensitive users toward other TEE‑hosted options. We read this as a strategic move to capture a middle market segment that values both reliability and blockchain‑native economics, potentially reshaping the competitive dynamics among centralized cloud providers and emerging decentralized AI inference platforms.
Sources used for this analysis
DeFi Intel wrote this analysis from the reporting above. Any figure or quotation can be checked against those pages.
Frequently asked questions
What model did NEAR AI Cloud add to its platform?
NEAR AI Cloud added Anthropic’s Claude Sonnet 5.5, a coding‑focused large language model with a one‑million‑token context window and adaptive reasoning enabled by default.
How is Claude Sonnet 5.5 priced on NEAR AI Cloud?
The model is priced at $2 per million input tokens and $10 per million output tokens, matching Anthropic’s public rates.
What is the performance of Claude Sonnet 5.5 on coding benchmarks?
Claude Sonnet 5.5 achieved a 70.6 % score on the Terminal‑Bench 4.0 coding benchmark, a significant improvement over its predecessor’s 10.3 % score.
DeFi Intel is an entity-graph aggregator: we curate, tag and link crypto news to a typed knowledge graph of protocols, tokens, people and incidents. We do not republish the full article body. Use the link above to read the original report at Crypto Briefing.
Discussion
Comments are reviewed before publication.
Want the full article?
Continue reading on Crypto Briefing →
No comments yet.