--- title: Together.ai Reviews meta_title: 'Together.ai Reviews November 2026: Details, Pricing, & Features | G2' meta_description: Filter 93 reviews by the user's company size, role or industry to find out how Together.ai works for a business like yours (as of November 2026). aggregate_rating: rating_value: 4.3 review_count: 93 scale: '5' date_modified: '2026-11-18' parent_category: name: Generative AI url: https://www.g2.com/categories/generative-ai ---

Together.ai Reviews & Product Details

Value at a Glance

Averages based on real user reviews.

Time to Implement

2 months

Return on Investment

8 months

User Insights

Average based on 93 real user reviews.

Implementation Time

2 months

Together.ai Reviews (93)

Reviews

Together.ai Reviews (93)

4.3
93 reviews

What do users say?

Generated using AI from real user reviews

Pros & Cons

Generated from real user reviews
View All Pros and Cons
Search reviews
Filter Reviews
Clear Results
G2 reviews are authentic and verified.
Yogesh S.
YS
Yogesh S.
Software Engineer
Small-Business (50 or fewer emp.)
"Fast open-source LLM hosting and OpenAI-compatible API integration"
5/5
What do you like best about Together.ai?

Together.ai provides an exceptionally fast, reliable, and cost-effective API infrastructure for running open-source LLMs. What stands out immediately is their OpenAI-compatible REST API interface and client SDK support. We were able to plug Together's endpoints directly into our existing backend microservices by changing just a few lines of configuration code. The streaming response performance is remarkably smooth with sub-100ms time-to-first-token (TTFT), making it ideal for powering customer-facing chat interfaces and automated backend tasks. Additionally, having access to hundreds of open-weights models (like Llama, Mistral, and Qwen) on a pay-as-you-go serverless model without managing complex GPU infrastructure gives us immense operational flexibility. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

While the serverless endpoint catalog is extensive, running specialized fine-tuned models or custom weights that aren't natively supported in the serverless pool requires spinning up dedicated GPU instances, which introduces higher fixed baseline costs. Furthermore, default rate limits and concurrent request limits on public serverless tiers can feel restrictive during sudden traffic spikes, requiring careful backend request queuing and retry logic unless you reserve dedicated capacity in advance. Lastly, while the web dashboard provides clear usage metrics, having deeper built-in request logging and tracing tools directly in the console would simplify day-to-day debugging. Review collected by and hosted on G2.com.

Ayush K.
AK
Ayush K.
Software Engineer
Mid-Market (51-1000 emp.)
"High-Speed Model Inference and Cost-Effective AI Scaling"
4.5/5
What do you like best about Together.ai?

The standout feature of Together.ai is its exceptional inference performance and ultra-low latency. It allows us to easily run, deploy, and scale open-source LLMs without the hassle of managing hardware cluster configurations. Its API integration is incredibly seamless, effectively functioning as a 'plug-and-play' alternative to OpenAI endpoints, which saved our engineering team weeks of setup time. Furthermore, features like 'dedicated model inference' and the recent support for 'multi-LoRA adapters' make fine-tuning and running customized workloads significantly more cost-effective compared to standard cloud GPU providers. It delivers enterprise-grade reliability and high throughput at a fraction of the cost. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

Although this platform delivers excellent overall performance, there are some areas where there is room for improvement. The available documentation for handling advanced configurations and specific edge cases is quite limited. Consequently, developers who lack extensive experience with model hosting might need to go through a fair amount of trial-and-error during the initial setup. Furthermore, during peak load, 'rate limits' can suddenly be triggered on serverless endpoints (unless you promptly upgrade to dedicated instances). Additionally, for cost monitoring, the dashboard metrics could provide more granular data on real-time usage. Review collected by and hosted on G2.com.

Shafiah S.
SS
Shafiah S.
Design Engineer
Mid-Market (51-1000 emp.)
"Fast, Affordable Inference with an Easy OpenAI SDK Swap"
3.5/5
What do you like best about Together.ai?

Honestly, the speed and price are hard to beat. Swapping out OpenAI's SDK for Together's to run Llama 3 or DeepSeek took me maybe ten minutes, and the response times are snappy enough that our users didn't even notice the switch. Getting sub-second inference without having to spin up, manage, or pay for expensive continuous GPU instances on AWS saves us a ton of money every month on side projects and smaller production apps. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

The main pain point is reliability during high traffic times. If you're using their shared serverless endpoints, you'll randomly hit rate limits or see latency spikes when a model is trending, forcing you to write retry logic or set up expensive dedicated endpoints. The developer dashboard also feels a bit basic tracking down granular API token usage per client or debugging failed requests could definitely be better. Review collected by and hosted on G2.com.

Raj C.
RC
Raj C.
Senior Software Engineer
Information Technology and Services
Small-Business (50 or fewer emp.)
Originally submitted 10/17/2026
"High-performance LLM API endpoints and low-latency inference"
5/5
What do you like best about Together.ai?

Together.ai is an exceptional choice for integrating open-source foundation models into production systems without managing low-level GPU infrastructure. Their serverless API endpoints deliver remarkably fast time-to-first-token (TTFT) and sustained generation throughput, making them ideal for latency-sensitive web and mobile applications. What makes Together.ai stand out is its OpenAI-compatible REST API and SDKs, which allow us to swap model backends (such as Llama 3, Mistral, or Qwen) with minimal code changes. Additionally, the ability to deploy dedicated instances for custom weights or fine-tuned LoRA adapters with predictable auto-scaling gives us high operational flexibility as application traffic grows. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

While the serverless endpoint pool covers all major open-source architectures, running non-standard or highly customized Transformer variants requires provisioning dedicated GPU instances, which increases baseline fixed costs. Furthermore, default rate limits and concurrency caps on public serverless endpoints can feel restrictive during sudden upstream traffic spikes unless you proactively reserve dedicated capacity. While basic API metrics and logs are available in the platform console, having more advanced built-in observability features—such as deep tracing for streaming response chunks—would simplify backend debugging. Review collected by and hosted on G2.com.

Kirpalsinh R.
KR
Kirpalsinh R.
Cyber Security Intern
Information Technology and Services
Small-Business (50 or fewer emp.)
"High-speed threat parsing and automated SIEM log enrichment API"
5/5
What do you like best about Together.ai?

Together.ai provides an exceptionally fast and cost-efficient platform for running open-source LLMs to analyze unstructured security events. Integrating Together.ai’s low-latency API endpoints into our security automation workflows allows us to pass raw PowerShell scripts, obfuscated command lines, and complex JSON logs to lightweight models like Llama 3 or Qwen for immediate parsing and intent summarization. The ultra-fast time-to-first-token (TTFT) and high token generation speeds ensure that log enrichment occurs in real time without creating queues in our triage pipeline. Additionally, the drop-in OpenAI API format allows us to seamlessly plug Together endpoints into our SOAR playbooks and custom Python enrichment scripts. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

When handling sensitive security logs and threat intelligence streams, verifying data retention policies across standard serverless endpoints requires careful oversight and configuration under enterprise agreements. Furthermore, during high-volume security incidents when SIEM alerts spike, default rate limits on serverless tiers can require setting up request queuing or spinning up dedicated GPU instances to avoid throttling. Adding more granular, security-focused compliance audit trails directly within the console dashboard would also enhance transparency for enterprise SOC auditing. Review collected by and hosted on G2.com.

Verified User in Marketing and Advertising
AM
Verified User in Marketing and Advertising
Small-Business (50 or fewer emp.)
"Together AI Makes It Easy to Experiment with Different AI Models"
5/5
What do you like best about Together.ai?

Together.ai is the flexibility it offers when experimenting with AI for business use cases. Since I work in AI marketing, I can see the value in testing different models for tasks like generating marketing content, analyzing information and building AI powered workflows. I like that it gives developers a way to explore different models without having to set up a separate integration for each one Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

It can take some time to get comfortable with if you're not deeply technical. Review collected by and hosted on G2.com.

Sanooj M.
SM
Sanooj M.
CEO
Small-Business (50 or fewer emp.)
Originally submitted 10/20/2026
"Fast, Affordable Single Platform for Open-Source Model Inference and Fine-Tuning"
4.5/5
What do you like best about Together.ai?

Fast, cheaper, single platform for all open source models for inference and finetuning. Rather than expensive options, this provides startups to get started faster with minimal capital. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

It has a good learning curve, but the platform isn’t very friendly for mom developers. The pricing is also a bit complex. That said, I think it’s cheap, and it’s almost the same as other similar platforms. Review collected by and hosted on G2.com.

Abhinav S.
AS
Abhinav S.
Security Operations Center Analyst
Information Technology and Services
Mid-Market (51-1000 emp.)
"Real-time SIEM log processing and threat parsing via LLM endpoints"
5/5
What do you like best about Together.ai?

Together.ai provides batter low-latency, scalable infrastructure for utilizing open-source LLMs to analyze and enrich complex security telemetry. By integrating Together.ai’s fast inference endpoints into our security automation pipelines, we can pass unstructured log streams, raw PowerShell scripts, and obfuscated shellcodes to lightweight models like Llama 3 or Qwen for instant parsing and intent extraction. The platform’s exceptionally low time-to-first-token (TTFT) and high token generation throughput mean that automated log triage happens in milliseconds. Furthermore, the OpenAI-compatible REST API allows us to easily drop Together endpoints into existing Python SOAR scripts and SIEM enrichment webhooks without rewriting our integration logic. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

Because security operations deal with sensitive internal telemetry and log data, ensuring strict data privacy and zero-data-retention compliance across public serverless endpoints requires careful review of enterprise tier settings. Additionally, when processing high-volume alert bursts from SIEM connectors, rate limits on standard serverless tiers can require setting up fallback request queues or provisioning dedicated GPU instances to avoid rate-limit throttling during major security events. Having built-in native security audit logging directly tailored for SOC compliance frameworks within the console would also be a welcome addition. Review collected by and hosted on G2.com.

MQ
Mahwish Q.
Software Engineer
Mid-Market (51-1000 emp.)
"Lightning-Fast, Affordable Inference—But Documentation and Scaling Can Be Tricky"
3.5/5
What do you like best about Together.ai?

Together.ai provides lightning-fast inference for leading open-source models like Llama, DeepSeek, and Mistral using a drop-in, OpenAI-compatible API. It removes the hassle of GPU management, drastically cutting down both development time and token costs. Though advanced fine-tuning docs could be clearer, it's unmatched speed and affordability make it ideal for scaling AI applications quickly. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

Together.ai has a steep learning curve for non-technical users, and its documentation for advanced workflows like custom deployments and fine-tuning can feel sparse. Developers may also encounter occasional rate limits during peak usage on high-demand serverless endpoints, and cost tracking gets complex when scaling to dedicated hardware. Review collected by and hosted on G2.com.

Shamshad B.
SB
Shamshad B.
Software Engineer
Computer Software
Small-Business (50 or fewer emp.)
"Fast, Low-Latency Serverless API That Makes Swapping Open-Source Models Easy"
4/5
What do you like best about Together.ai?

What stands out most is the interface speed and ease to integration. Being able to query leading open-source models (like Llama, Mistral, and Qwen) via an OpenAI-compatible API allows us to swap models into our existing stack with minimal effort. The serverless endpoint latencies are consistently low, and avoiding the hassle of managing dedicated GPU hosting for prototyping and production workloads saves a ton of engineering overhead. Review collected by and hosted on G2.com.

What do you dislike about Together.ai?

While the core API performance is great, the platform documentation can occasionally feel thin around specific edge, like custom LoRA deployments or detailed rate limit handling. Additionally, because there are so many open-source model variants hosted on the platform, navigating model availability changes or occasional endpoint depreciations requires keeping a close eye on your integration code. Review collected by and hosted on G2.com.

Questions about Together.ai? Ask real users or explore answers from the community

Get practical answers, real workflows, and honest pros and cons from the G2 community or share your insights.

Sanjeev M.
SM
Sanjeev Mishra
•
Last activity 20 days ago

How beginner-friendly do you find Together.ai pricing and documentation for getting started?

Pricing Insights

Averages based on real user reviews.

Time to Implement

2 months

Return on Investment

8 months

Together.ai Features
AI Cost per API Call
AI Resource Allocation Flexibility
AI Energy Efficiency
Fine-Tuning Interface
AI Data Pipeline Integration
AI API Support and Flexibility
One-Click Deployment
Scalability Management