Together AI

What do you like best about the product?

What I like best about Together.ai is how it provides access to different AI models in one platform. I appreciate the flexibility of being able to explore different models depending on the task instead of being limited to just one. The platform is also relatively straightforward to use once you understand the basics, and its API integrations make it convenient to incorporate AI into existing workflows.

Another thing I like is the performance. Responses are generally fast, and the platform offers options for balancing speed, quality, and cost. Overall, it's a useful tool for experimenting with AI and improving productivity.

What do you dislike about the product?

One thing I dislike about Together.ai is that it can be a little overwhelming for beginners, especially with the number of models and configuration options available. It takes some time to understand which models work best for specific tasks, and the documentation can sometimes feel too technical for users without a development background.

I also think monitoring usage and costs could be more straightforward. While the pricing is flexible, expenses can add up depending on the models being used and the volume of requests. Overall, it's a powerful platform, but a simpler onboarding experience would make it more accessible.

What problems is the product solving and how is that benefiting you?

Together.ai helps solve the challenge of accessing and testing different AI models without having to build and maintain the infrastructure yourself. Instead of spending time setting up models individually, the platform provides a centralized environment where you can experiment with different AI capabilities.

The biggest benefit is the flexibility and time savings. It makes it easier to compare model performance, automate repetitive tasks, and explore AI-powered solutions for different workflows. This can help improve productivity and reduce the technical effort required to implement AI tools.

What do you like best about the product?

Together makes it really convenient for me to experiment with different language models and incorporate inference APIs into my software products. I use it for prompt testing, response comparisons, and latency assessments for my choice of the underlying models. The API interaction is simple, and the wide model range allows me to optimize for price, latency, and output quality.

What do you dislike about the product?

Maybe the dashboard can help users better understand advanced model configurations and price comparisons. Performance on the models can vary so it would also be nice if more testing was done before deployment. I would have liked a more detailed walkthrough and faster support when troubleshooting API errors or unexpected inferences.

What problems is the product solving and how is that benefiting you?

Together AI minimizes the engineering required to test and deploy models within my application. I can compare inference engines, review responses, and choose models by latency, accuracy, and cost. As an API, Together AI diminishes the amount of custom integration work required to get up and running, enabling me to prototype quickly and devote more time to code, test, and production.

What do you like best about the product?

I have liked Together.ai for how easy it is to try different AI models from one place. We have been experimenting with a few different models and having the flexibility to switch between them without setting everything up from scratch has been useful. The API was also fairly easy to understand and get started with. For me, the biggest benefit is being able to test different models and see which one works better for a particular requirement.

What do you dislike about the product?

The number of models available can sometimes make the choice a little confusing. It took me some time to figure out which models were actually worth testing for our use case.. I think some clearer comparisons or recommendations based on different use cases would make it easier, especially for someone who is just getting started.

What problems is the product solving and how is that benefiting you?

It mainly helps us with the testing and experimentation side of AI. Instead of spending a lot of time integrating different models separately, we can try different options and compare the results. That has made it easier for us to decide which model make sense before putting more effort into an implementation.

What do you like best about the product?

Straight forward, user friendly in formatting and guiding aids with quick response time

What do you dislike about the product?

Takes a little time with training the technology to direct what you actually want, meaning if I formulate a task, it is something I could have googled myself instead of Together.ai computing their offerings.

What problems is the product solving and how is that benefiting you?

Solving projecting prospects aligning with organizational needs, searching data in past to present years of funding from all organizational levels, from funding to future funding priorities.

What do you like best about the product?

We use the Together Inference engine and their dedicated GPU clusters every day in our production pipelines. As a result, token latency for our Llama models has dropped by 40%, and our developers can now deploy custom models in minutes rather than hours. The dashboard UI is intuitive and responsive, and the OpenAI-compatible SDK and API integration were straightforward enough that they cut onboarding time for our infrastructure team in half. During testing, we were also pleasantly surprised that the built-in quantization options and fine-tuning workflows reduced our system’s memory consumption by about 30%, with no loss in accuracy

What do you dislike about the product?

During peak hours, when usage spikes unexpectedly, I occasionally notice a slight delay before the dashboard shows detailed cost analytics and usage metrics. If the team could improve the speed of these near-real-time reporting updates, it would make it even easier for us to track our monthly cloud budget internally.

What problems is the product solving and how is that benefiting you?

We previously struggled to manage heavy GPU instances on hyperscaler clouds while also dealing with the high infrastructure costs that came with them. Now we can rely on simple serverless API calls and scale instantly based on demand. As a result, we’ve seen roughly 45% direct savings on our monthly infrastructure bill and a noticeably better ROI.

Since making this transition, our cloud operations have been completely free of manual cluster maintenance, which saves the team at least 12–15 hours of valuable engineering time every week. We also no longer have to wait weeks to test newly released open-weight models in production. Overall, this has led to meaningful improvements in both system reliability and deployment speed.

What do you like best about the product?

I use Together AI for trying out open source language models, testing out inference results and adding AI features into prototyping applications. The playground allows for easy experimentations and the API doc helps with integration. I like the model choices, versatile metered pricing, and integrations like LangChain that help you get started for various needs.

What do you dislike about the product?

A dashboard would be nice for watching usages or comparing models. Response times are not always predictable, depending on model complexity and workload, so I'd like to see some more performance testing. I'd like some more troubleshooting examples and diagnoses for integration problems that don't involve trial and error.

What problems is the product solving and how is that benefiting you?

Together AI simplifies the infrastructure overhead of testing and deploying models into applications. I can compare model outputs, assess quality, and pick the right option without handling the full inference stack. AI-assisted development workflows enable more rapid prototyping, and usage-based pricing allows me to keep costs manageable and spend more time on features.

What do you like best about the product?

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.

What do you dislike about the product?

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.

What problems is the product solving and how is that benefiting you?

Together.ai eliminates the time-consuming process of manually dissecting complex, obfuscated threat payloads and unformatted security logs. Before incorporating Together.ai endpoints into our automated triage steps, analysts had to spend valuable time manually decoding base64 strings, analyzing suspicious scripts, and assembling incident context. Together.ai automates this pre-triage intelligence gathering, delivering clear, summarized threat context directly to our incident tickets before an analyst opens them, which drastically lowers our Mean Time to Respond (MTTR).

What do you like best about the product?

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

What do you dislike about the product?

It can take some time to get comfortable with if you're not deeply technical.

What problems is the product solving and how is that benefiting you?

One thing I find useful about Together AI is that I can try different AI models without having to jump between multiple platforms. Since I work in AI mar-tech, I’m always curious about how we can use AI to improve content, automate repetitive tasks or solve everyday business problems. It gives me a way to experiment with different models and see what works best without having to build everything from scratch

What do you like best about the product?

What I like best about Together.ai is its fast inference speed and reliable, high-performance API access to leading open-source models. It makes running and deploying open models seamless and cost-effective.

What do you dislike about the product?

Occasionally, rate limits can be restrictive during peak times, and high-demand open-source models sometimes experience temporary unavailability or slowdowns.

What problems is the product solving and how is that benefiting you?

Together.ai eliminates the friction of prototyping with open-source models. It allows me to test, fine-tune, and switch between different open-source architectures instantly through a unified API, drastically reducing time-to-market for AI-powered features.

What do you like best about the product?

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.

What do you dislike about the product?

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.

What problems is the product solving and how is that benefiting you?

Together.ai solves the bottleneck of manually analyzing complex, unstructured security events and malicious payloads. Prior to leveraging Together.ai endpoints in our triage workflows, analysts spent significant time manually decoding base64 commands, dissecting suspicious script blocks, and summarizing raw incident logs. Together.ai allows us to run cost-effective, high-throughput LLM analysis on security events automatically before an analyst opens the ticket, drastically reducing MTTR (Mean Time to Respond) and cutting down manual investigative overhead.