The Russian banking and technology group's latest model combines specialised reasoning with a 262,000-token context window as businesses assess AI for software, analytics and operational workflows.
Sber has introduced GigaChat 3.5 Reasoning, an artificial intelligence model designed to plan, verify and work through complex, multi-stage tasks. The launch expands the group's enterprise AI proposition as organisations look beyond conversational tools towards systems capable of completing connected business processes.
Announced on September 10, 2026, the model can divide problems into stages, identify missing information, use external tools and check intermediate results before responding, according to the company. It can also revisit earlier decisions when a proposed solution does not satisfy the task's requirements.
Sber said the model recorded its largest improvements over GigaChat 3.5 Ultra Instruct in mathematics, coding, instruction following and structured output.
Enterprise use cases and business value
The model is aimed at tasks where the quality of an answer depends on several connected decisions. These include debugging software, examining documents for conflicting provisions, preparing schedules within budget and resource constraints, and handling customer requests that require information from multiple systems.
For businesses, the potential value lies in reducing the manual coordination between these steps. Software development is one of its primary applications. The model has been trained on algorithms, code editing, test generation and repository-level assignments in which proposed patches are applied and tested.
In document analysis, the model can compare requirements across contracts and identify potentially inconsistent clauses for further review. Planning applications could include building routes or schedules while accounting for deadlines, budgets and available resources.
Customer-support systems could use the model to clarify requests and consult different sources before producing a response. Analytical workflows could similarly combine calculations with successive checks of a hypothesis. For companies evaluating deployment, returns will depend on integration quality, employee oversight and whether improvements in accuracy offset additional computing and implementation costs.
Anton Frolov, senior vice president and head of GenAI Development at Sberbank, said the ability to identify missing information, validate intermediate conclusions and modify a proposed solution is particularly important in programming, analytics and agent-based applications.
Specialised training for complex tasks
GigaChat 3.5 Reasoning is the first model in the GigaChat family to feature full reasoning trained through online reinforcement learning, according to its technical documentation.
Its post-training process uses six independently trained domain experts covering science and mathematics, coding, code agents, general agents, dialogue, and capabilities including instruction following, long-context processing and structured output.
Different evaluation methods are used for each area. Mathematical tasks are assessed through final-answer verification, coding exercises through code execution and repository-level changes through tests applied after a patch. General agent tasks are measured against the resulting state of the environment, while dialogue responses undergo comparative assessment by an AI judge.
The experts are subsequently combined through on-policy distillation. Under this approach, the main model generates its own solution path while the relevant expert provides token-level supervision.
Scale performance and deployment economics
The model uses a Mixture-of-Experts architecture containing 432 billion parameters, of which 28 billion are active during processing. It supports a maximum context length of 262,000 tokens, allowing it to work with substantial documents, extended conversations and large collections of code.
Its hybrid architecture combines Multi-head Latent Attention with GatedDeltaNet linear-attention layers. The model was trained natively in FP8 precision, while a BF16 version is available for fine-tuning and custom quantisation.
Sber reported that the model's IFBench score increased from 44 to 77 compared with the non-reasoning version. Its Natural Plan score rose from 64 to 80, while its Live Code Bench v6 score increased from 56 to 85.
The company also claimed that the model used an average of 37% fewer tokens on mathematical problems than DeepSeek V4 Flash Preview. The comparison relates to a specific workload and does not necessarily translate into an equivalent reduction in operating costs or response times.
Users can activate reasoning mode for more complex requests, while straightforward questions can receive immediate responses. Sber said difficult problems may require tens of thousands of tokens, leaving enterprises to balance accuracy against processing time and resource consumption.
GigaChat 3.5 Reasoning is available to users and developers, with its model weights released on Hugging Face under the MIT licence. Sber has also said that API access for businesses will follow. For Sber, the commercial opportunity will emerge as companies move from internal trials to governed, repeatable deployments.
The release gives enterprises another open-weight option to test against proprietary AI systems. Its wider adoption will depend on whether organisations can convert its technical capabilities into reliable workflows that meet requirements for accuracy, data governance, speed and cost.
NOTE: This article is created by VCCEdge Research team.
Insights Focus is a marketing initiative for sponsored posts. No VCCircle journalist was involved in the creation of this content.